View Issue Details
466 [JANA] Bug minor always 2015-10-26 14:19 2016-12-08 11:03
davidl  
davidl  
normal  
resolved  
fixed  
none    
none  
   
JResourceManager doesn't test path from jcalib
This problem came up when trying to run using the resources directory in /group/halld. The issues is that the Magnets/Solenoid/finemeshes/ directory contains files that have a ".evio" suffix. The namepaths corresponding to these files do not have this suffix. This would normally be OK if the files were obtained through the JResourceManager since it would write an entry in the top-level "resources" file that would contain the ".evio" suffix and therefore, return the full path to this file. However, the resources directory in /group/halld has no "resources" file. This means GetResource looks for a file based only on the namepath (i.e. without ".evio") and fails to find it. It then tries downloading it, but fails since it can't write to the resources directory from a personal account.

To get around this, GetResource should check for a filename whose path comes from jcalib before attempting a download. This should probably happen only if it fails to find a file based on the namepath (around line 230 in JResourceManager.cc).
Notes
(0000656)
davidl   
2015-11-29 15:56   
OK, it's not so clear now what was going on here. The default behavior of GetResource is to call GetLocalPathToResource() which will use the jcalib object to get the full path based on the name in CCDB. It should only return the full path based on the name path if no jcalib exists. This should not have been the case before, but I cannot recall now the exact conditions leading to the issue. I'm going to have to defer fixing this until the problem presents again.
(0000666)
davidl   
2016-12-08 11:03   
This looks to have been fixed back in March 2016 and included in jana 0.7.5 (see svn entry 2412)





View Issue Details
471 [JANA] Bug minor always 2016-03-06 09:13 2016-03-11 15:35
davidl  
davidl  
normal  
resolved  
fixed  
none    
none  
   
janaroot variable names containing colons
The toStrings mechanism in JANA uses colons to as a delimiter to separate columns. This conflicts with our EPICS variable names which also use colons. The code that determines the string that is the name of the column in the janaroot plugin needs to use the remainder of the string and not just tokens[2] as the value.
Notes
(0000665)
davidl   
2016-03-11 15:35   
Fixed in revision 2410 and will appear in JANA 0.7.5





View Issue Details
470 [JANA] Bug minor always 2016-01-29 15:24 2016-03-11 10:57
davidl  
davidl  
normal  
resolved  
fixed  
none    
none  
   
-factoryreport crashes when factory name too long
there is no test in JApplication::PrintFactoryReport to check that a string passed into string::replace is not bigger than the original. This results in an application-killing exception being thrown.
Notes
(0000664)
davidl   
2016-03-11 10:57   
This was fixed in revision 2409 and will appear in JANA 0.7.5





View Issue Details
451 [JANA] Feature Request minor N/A 2015-02-12 13:43 2015-11-30 11:51
davidl  
davidl  
normal  
resolved  
fixed  
none    
none  
   
Add scripts/macro to make standard rate plots to JANA source
Periodically benchmarking needs to be done using JANA or JANA-based code. I seem to reinvent the wheel on this a lot so it would be good to add a script or two along with ROOT macros that could be easily used to make some standard plots.
Notes
(0000662)
davidl   
2015-11-30 11:51   
Added macro directory in revision 2379 and will appear in JANA 0.7.4





View Issue Details
450 [JANA] Feature Request minor N/A 2015-02-04 13:54 2015-11-30 10:16
davidl  
davidl  
normal  
resolved  
fixed  
none    
none  
   
Need SKIP_TO_EVENT config. parameter
Suggested by Matt S.:

We should add a SKIP_TO_EVENT configuration parameter that would allow one to specify the event number (as reported by the JEvent object) that should be skipped to. This is somewhat related to the function of EventStore that provides a list of events to process. We expect to implement the EventStore at the JEventSource_EVIO level (i.e. not as part of JANA, but as part of sim-recon). However, it may be easy to add the ability to use EventStore formatted files.
Notes
(0000661)
davidl   
2015-11-30 10:16   
Looks like this was already implemented back on Jun 10th, 2015 and came out in JANA 0.7.3 already.





View Issue Details
467 [JANA] Feature Request minor N/A 2015-11-02 09:17 2015-11-30 10:09
davidl  
davidl  
normal  
resolved  
fixed  
none    
none  
   
Default JANA_RESOURCE_DIR
This is based on a suggestion by Mark I. after and e-mail conversation that came up due to a lot of farm jobs simultaneously hitting the web server for resources during start up. I can't seem to find Mark's e-mail so this is from memory.

We should have the ability to specify a default value for the JANA_RESOURCE_DIR in the sim-recon code in case the user does not set JANA_RESOURCE_DIR in their environment. This can point to the directory on /group/halld in the case of sim-recon. This directory can then be checked first before falling to the /tmp option. This will be a little subtle since the default behavior will change depending on whether you are running on a machine with /group/halld mounted our not. However, most of the time, it will be doing the right thing. This also puts the responsibility for using this default feature into sim-recon and not JANA itself so that seems OK.
Notes
(0000660)
davidl   
2015-11-30 10:09   
Fixed in revision 2377. This will appear in JANA 0.7.4

The fix was to add a new config. param called JANA_RESOURCE_DEFAULT_PATH. This is a colon separated list of directories to look for before falling to the last resort of the /tmp directory. The first directory in the list found to exist is used. This will allow us to specify more than one default. The idea being that different systems may have different defaults that are unlikely to exist on other systems.





View Issue Details
460 [JANA] Feature Request minor have not tried 2015-06-10 10:32 2015-11-30 07:54
davidl  
davidl  
normal  
resolved  
fixed  
none    
none  
   
Remove "int" version of JEventProcessor::evnt
The int version of JEventProcessor::evnt needs to be removed and the uint64_t version not made to call it. This should only be done after JANA 0.7.3 is commonplace.
Notes
(0000659)
davidl   
2015-11-30 07:54   
Fixed in revision 2376 and will appear in JANA 0.7.4. This required changes to many files including the helper scripts "mkXXX".





View Issue Details
463 [JANA] Feature Request minor N/A 2015-08-10 07:35 2015-11-30 06:40
davidl  
davidl  
normal  
resolved  
fixed  
none    
none  
   
ROOT 6 support
Instances of __CINT__ must also include __CLING__ in order to support ROOT 6. See e-mail from Nathan Sparks on July 23, 2015.
Notes
(0000658)
davidl   
2015-11-30 06:40   
Fixed in revision 2375 and will appear in JANA 0.7.4





View Issue Details
464 [JANA] Feature Request minor N/A 2015-08-10 07:43 2015-11-30 06:15
davidl  
davidl  
normal  
resolved  
fixed  
none    
none  
   
Use c++,cc instead of g++,gcc
Modify the SConstruct file to use c++,cc instead of g++, gcc so that one can easily use alternative compilers such as clang. Don't forget to pull in shell environment so invoking shell's PATH and LD_LIBRARY_PATH are used.
Notes
(0000657)
davidl   
2015-11-30 06:15   
Fixed in revision 2374 and will appear in JANA 0.7.4





View Issue Details
468 [JANA] Bug minor always 2015-11-06 14:40 2015-11-29 15:27
davidl  
davidl  
normal  
resolved  
fixed  
none    
none  
   
DEFTAG: feature does not work
Simon reported and I confirmed that the DEFTAG: feature in JANA stopped functioning in JANA 0.7.3. This was due to a line being added the JEventLoop::Initialize() that tries to get the AUTOACTIVATE parameter just before getting ll of the DEFTAG: parameters. This is done inside a try-catch block (line 426). If the AUTOACTIVATE parameters is not set, then an exception is thrown and the DEFTAG: call is not made. As a result, one can actually get the DEFTAG: feature to work again if the AUTOACTIVATE parameter is set to something.

The fix is to move each of the GetParameter calls into its own try-catch block so if one fails, the others still get tried.
Notes
(0000655)
davidl   
2015-11-29 15:27   
Fixed in revision 2373 and will appear in JANA 0.7.4





View Issue Details
469 [JANA] Bug minor N/A 2015-11-18 08:45 2015-11-29 15:18
davidl  
davidl  
normal  
resolved  
fixed  
none    
none  
   
Examples missing SConstruct/SConscript
The two examples are missing SConstruct/SConscript files. The Makefiles are no longer there either and the README files refer to them. An SConscript file should be added to each and the examples made into Optionally-Build items, The README files need to be updated so reflect this.
Notes
(0000654)
davidl   
2015-11-29 15:18   
Submitted in revision 2371 and will be available in JANA 0.7.4





View Issue Details
465 [JANA] Bug minor N/A 2015-10-20 15:52 2015-11-29 15:18
davidl  
davidl  
normal  
resolved  
fixed  
none    
none  
   
JANA program does not exit with non-zero code
This was brought up by Paul M. regarding some CLAS6 jobs he was running. A few of his jobs reached a thread time-out causing JANA to quit early. Without the relaunch flag set, this will print a message and set the Quit() flag. JANA then exits cleanly (i.e. returns from Run() ) and the program doesn't know to return a non-zero exit value. The farm manager then indicates that the job was a SUCCESS, even though it really wasn't.

We don't necessarily want a thread time-out to cause an immediate exit. At the same time, though, we do want the program to exit with a non-zero code so that the farm manager can detect this. Options are:

1. If in BATCH mode, do exit immediately

2. set an exit code in JApplication that should be returned from Run() and may also be accessible using the JApplication pointer. This will allow the actual program (e.g. hb_root) to check for this and return an appropriate exit code.


Notes
(0000650)
kmoriya   
2015-10-27 10:15   
For farm jobs, Kei did tests with
exit $exit_signal
and found that both SWIF and Auger are able to pick up this exit code.
SWIF will also mark the job with problem tag "SWIF-USER-NON-ZERO".

Exit codes higher than 128 and in the range (30,40) should be avoided
since Chris uses these within SWIF.
(0000651)
davidl   
2015-11-24 16:47   
Kei also noted that when a file doesn't exist and an error message is printed, the exit code is still "0". This should be set to something non-zero.
(0000653)
davidl   
2015-11-29 15:18   
Fixed in revision 2371 and will appear in JANA 0.7.4.

JANA utilities and examples updated to use it and provide example on how to.





View Issue Details
418 [JANA] Feature Request minor N/A 2014-09-29 07:33 2015-11-29 14:15
davidl  
davidl  
normal  
resolved  
fixed  
none    
none  
   
Add -a option to janadump
hd_dump has a "-a" option that will cause associated objects to be printed. This should be added to janadump as well. An improvement would be to allow a numerical value to be specified indicating the depth of associations to go so that one could print out not only the directly associated objects, but the associated objects of those associated objects as well.
Notes
(0000652)
davidl   
2015-11-29 14:15   
Committed in revision 2370 and will appear in JANA 0.7.4.

(n.b. the numerical depth option wasn't actually added)





View Issue Details
462 [Hall D Offline] General minor N/A 2015-06-15 10:25 2015-06-15 10:25
davidl  
 
normal  
new  
open  
none    
none  
   
hddm_merge_events broken
After the major commit in svn revision 14984 that upgraded everything to use the C++ API for HDDM, the hddm_merge_events program was broken. The problem is that upgrade left it writing out each event as a separate HDDM event in the output file rather than adding it as multiple PhysicsEvent's. Although it is likely that the original intent of the PhysicsEvent structure in HDDM was to house separate events, the JEventSource_HDDM class in DANA loops over these, effectively merging the events (sans proper handling of overlapping hits in a single channel.)
There are no notes attached to this issue.





View Issue Details
461 [Hall D Offline] General minor have not tried 2015-06-10 10:37 2015-06-10 10:37
davidl  
 
normal  
new  
open  
none    
none  
   
Update all processors and factories to use uint64_t for event number
The JANA API has been modified in version 0.7.3 to use uint64_t for event numbers. A wrapper method was included in JEventProcessor though to allow for backward compatibility. One JANA 0.7.3 is commonplace, the switch should be made in sim-recon to prepare for the eventual removal of the deprecated "int" version.
There are no notes attached to this issue.





View Issue Details
459 [Hall D Offline] General minor N/A 2015-06-10 10:26 2015-06-10 10:26
davidl  
 
normal  
new  
open  
none    
none  
   
Remove DCalibrationCCDB from DANA library
The DCalibrationCCDB class and its associated generator class overlap with the JCalibrationCCDB class and its generator in the JANA library. Differences were merged into JANA 0.7.3 to that we can drop DCalibrationCCDB altogether from sim-recon. This shouldn't happen though until 0.7.3 becomes commonplace so as not to require immediate upgrades for everyone.
There are no notes attached to this issue.





View Issue Details
3 [Hall D Offline] General minor have not tried 2010-06-18 09:12 2015-06-10 10:05
wolin  
marki  
normal  
resolved  
fixed  
none    
none  
   
Look into restoring doxygen generation and subversion statistics generation to the nightly build
From Wiki action item page
Notes
(0000038)
marki   
2010-09-16 16:07   
Reminder sent to: davidl
What are the next steps on this issue? Or are we done?
(0000042)
marki   
2010-10-06 16:44   
Marking as resolved. Both functions are part of nightly build now.





View Issue Details
51 [Hall D Offline] General crash always 2011-03-21 07:45 2015-06-10 10:05
jonesrt  
davidl  
high  
resolved  
fixed  
none    
none  
   
standard reconstruction segfaults in GetLastDOCAPoint()
Trying to complete a reconstruction sample on the grid, and getting a recurring segfault in reconstruction. How to reproduce:

1. Grab the file http://zeus.phys.uconn.edu/halld/gridwork/bug-3-21-2011/events_3806.hddm [^] - a single event that causes the segfault.
2. Build the latest version of the reconstruction code (as of 3/18/2011) on either x86_64 or i686 platforms, both show the same feature. I have only tested this with default optimization -O2.
3. Run "hd_ana -PPLUGINS=danahddm -PTHREAD_TIMEOUT=300 events_3806.hddm".
4. Watch it segfault.
Notes
(0000065)
davidl   
2011-03-21 10:40   
Was not able to reproduce this problem on OS X, but was on CentOS 5.3 (64bit).

Committed fix in svn revision 7603. It was due to use of a NULL pointer for last_swim_step in DReferenceTrajectory::GetLastDOCAPoint.

Currently awaiting confirmation of fix from Richard before marking issue as resolved.
(0000066)
jonesrt   
2011-03-21 14:41   
David's fix resolves this problem for me, confirm that this ticket can be closed.
(0000067)
davidl   
2011-03-21 14:44   
Fix has been confirmed. Marking issue as resolved.





View Issue Details
86 [Hall D Offline] General tweak have not tried 2011-06-02 12:01 2015-06-10 10:05
marki  
kornicer  
normal  
assigned  
open  
none    
none  
   
Decide on cluster merging after or before charged track matching
as discussed at the June 1, 2011 Offline Meeting
Notes
(0000284)
marki   
2012-01-31 16:48   
Reminder sent to: davidl, kornicer
This issue is resolved, right?





View Issue Details
456 [Hall D Offline] General trivial always 2015-03-17 13:12 2015-06-10 09:59
kmoriya  
davidl Mac OSX  
low 10.10 (Yosemite)  
resolved  
fixed  
none    
none  
   
osrelease.pl needs update
osrelease.pl, contained in BMS/ needs the following lines
around line 96 to accomodate OSX 10.10 (Yosemite):

 elsif ($release_string =~ /^14.*/) {
        $release = '_macosx10.10';

This applies also to other osrelease.pl files.
Notes
(0000648)
davidl   
2015-06-10 09:59   
This has now been done. It turns out there are at least 4 copies of this script in our code base:

scripts/osrelease.pl
scripts/build_scripts/osrelease.pl
sim-recon/src/BMS/osrelease.pl
$JANA_HOME/src/../SBMS/osrelease.pl

They have not been kept in strict alignment so after updating the one on build_scripts, I copied to the other 3 locations and committed it there as well.





View Issue Details
424 [Hall D Offline] General crash have not tried 2014-12-02 16:52 2015-06-10 09:42
davidl  
davidl  
normal  
resolved  
fixed  
none    
none  
   
Seg fault from mcsmear
A seg. fault is reported when running mcsmear using conditions described in Tom's e-mail copied in the "Additional Information" section. In a subsequent e-mail, he notes that he's using Linux_CentOS6-x86_64-gcc4.4.6 and that files demonstrating the problem can be found here:

https://docs.google.com/a/umail.iu.edu/file/d/0B2d57qMzynA9ejQ5Q3I0TDNTelE/edit?usp=drive_web [^]
Hi all,

I’m trying to generate some MC using the commissioning branch and simulation conditions. I’m generating some events using genr8 and running them through hdgeant using control.in_9101. When I try to run mcsmear on the output, I get a segfault after processing a couple hundred events. I see the same behavior if I generate events using bggen. I included a stack trace below. Does anyone have any advice for fixing this?

Tom Boettcher

Program received signal SIGSEGV, Segmentation fault.
[Switching to Thread 0x7ffff15f2700 (LWP 10822)]
0x0000000000bec50e in sgetn (this=0x7fffec5e9190, s="")
   at /usr/lib/gcc/x86_64-redhat-linux/4.4.6/../../../../include/c++/4.4.6/streambuf:335
335 { return this->xsgetn(__s, __n); }
Missing separate debuginfos, use: debuginfo-install bzip2-libs-1.0.5-7.el6_0.x86_64 cyrus-sasl-lib-2.1.23-13.el6_3.1.x86_64 freetype-2.3.11-14.el6_3.1.x86_64 glibc-2.12-1.80.el6_3.7.x86_64 keyutils-libs-1.4-4.el6.x86_64 krb5-libs-1.9-33.el6_3.3.x86_64 libcom_err-1.41.12-12.el6.x86_64 libcurl-7.19.7-26.el6_2.4.x86_64 libgcc-4.4.6-4.el6.x86_64 libidn-1.18-2.el6.x86_64 libselinux-2.0.94-5.3.el6.x86_64 libssh2-1.2.2-11.el6_3.x86_64 libstdc++-4.4.6-4.el6.x86_64 nspr-4.9.2-0.el6_3.1.x86_64 nss-3.13.6-2.el6_3.x86_64 nss-softokn-freebl-3.12.9-11.el6.x86_64 nss-util-3.13.6-1.el6_3.x86_64 openldap-2.4.23-26.el6_3.2.x86_64 openssl-1.0.1e-30.el6_5.2.x86_64 pcre-7.8-4.el6.x86_64 zlib-1.2.3-27.el6.x86_64
(gdb) where
#0 0x0000000000bec50e in sgetn (this=0x7fffec5e9190, s="")
   at /usr/lib/gcc/x86_64-redhat-linux/4.4.6/../../../../include/c++/4.4.6/streambuf:335
0000001 xstream::xdr::istream::operator>> (this=0x7fffec5e9190, s="") at external/xstream/src/xdr.cpp:133
0000002 0x00000000005bd5dd in hddm_s::BcalSiPMSpectrum::streamer (this=0x7fffec882290, istr=...)
   at /home/fs7/tboettch/HallD_Home/install/sim-recon/sim-recon-commissioning/Linux_CentOS6-x86_64-gcc4.4.6/include/HDDM/hddm_s.hpp:16529
0000003 0x00000000005bce25 in hddm_s::istream::sequencer (this=0x7fffec000d10, object=...)
   at /home/fs7/tboettch/HallD_Home/install/sim-recon/sim-recon-commissioning/Linux_CentOS6-x86_64-gcc4.4.6/include/HDDM/hddm_s.hpp:4961
0000004 0x00000000005e4ab1 in hddm_s::HDDM_ElementList<hddm_s::BcalSiPMSpectrum>::streamer (this=0x7fffec75a308, istr=...)
   at /home/fs7/tboettch/HallD_Home/install/sim-recon/sim-recon-commissioning/Linux_CentOS6-x86_64-gcc4.4.6/include/HDDM/hddm_s.hpp:428
0000005 0x00000000005bce25 in hddm_s::istream::sequencer (this=0x7fffec000d10, object=...)
   at /home/fs7/tboettch/HallD_Home/install/sim-recon/sim-recon-commissioning/Linux_CentOS6-x86_64-gcc4.4.6/include/HDDM/hddm_s.hpp:4961
0000006 0x00000000005bccaf in hddm_s::istream::operator>> (this=0x7fffec000d10, object=...)
   at /home/fs7/tboettch/HallD_Home/install/sim-recon/sim-recon-commissioning/Linux_CentOS6-x86_64-gcc4.4.6/include/HDDM/hddm_s.hpp:4951
0000007 0x00000000005bcec7 in hddm_s::istream::sequencer (this=0x7fffec000d10, object=<value optimized out>)
   at /home/fs7/tboettch/HallD_Home/install/sim-recon/sim-recon-commissioning/Linux_CentOS6-x86_64-gcc4.4.6/include/HDDM/hddm_s.hpp:4973
0000008 0x00000000005e5140 in hddm_s::HDDM_ElementList<hddm_s::BcalCell>::streamer (this=<value optimized out>, istr=...)
   at /home/fs7/tboettch/HallD_Home/install/sim-recon/sim-recon-commissioning/Linux_CentOS6-x86_64-gcc4.4.6/include/HDDM/hddm_s.hpp:428
0000009 0x00000000005bce25 in hddm_s::istream::sequencer (this=0x7fffec000d10, object=...)
   at /home/fs7/tboettch/HallD_Home/install/sim-recon/sim-recon-commissioning/Linux_CentOS6-x86_64-gcc4.4.6/include/HDDM/hddm_s.hpp:4961
0000010 0x00000000005bccaf in hddm_s::istream::operator>> (this=0x7fffec000d10, object=...)
   at /home/fs7/tboettch/HallD_Home/install/sim-recon/sim-recon-commissioning/Linux_CentOS6-x86_64-gcc4.4.6/include/HDDM/hddm_s.hpp:4951
0000011 0x00000000005bcec7 in hddm_s::istream::sequencer (this=0x7fffec000d10, object=<value optimized out>)
   at /home/fs7/tboettch/HallD_Home/install/sim-recon/sim-recon-commissioning/Linux_CentOS6-x86_64-gcc4.4.6/include/HDDM/hddm_s.hpp:4973
#12 0x00000000005bccaf in hddm_s::istream::operator>> (this=0x7fffec000d10, object=...)
   at /home/fs7/tboettch/HallD_Home/install/sim-recon/sim-recon-commissioning/Linux_CentOS6-x86_64-gcc4.4.6/include/HDDM/hddm_s.hpp:4951
0000013 0x00000000005bcec7 in hddm_s::istream::sequencer (this=0x7fffec000d10, object=<value optimized out>)
   at /home/fs7/tboettch/HallD_Home/install/sim-recon/sim-recon-commissioning/Linux_CentOS6-x86_64-gcc4.4.6/include/HDDM/hddm_s.hpp:4973
0000014 0x00000000005bccaf in hddm_s::istream::operator>> (this=0x7fffec000d10, object=...)
   at /home/fs7/tboettch/HallD_Home/install/sim-recon/sim-recon-commissioning/Linux_CentOS6-x86_64-gcc4.4.6/include/HDDM/hddm_s.hpp:4951
0000015 0x00000000005bcec7 in hddm_s::istream::sequencer (this=0x7fffec000d10, object=<value optimized out>)
   at /home/fs7/tboettch/HallD_Home/install/sim-recon/sim-recon-commissioning/Linux_CentOS6-x86_64-gcc4.4.6/include/HDDM/hddm_s.hpp:4973
0000016 0x00000000005dcb5f in hddm_s::HDDM_ElementList<hddm_s::PhysicsEvent>::streamer (this=<value optimized out>,
   istr=...)
---Type <return> to continue, or q <return> to quit---
   at /home/fs7/tboettch/HallD_Home/install/sim-recon/sim-recon-commissioning/Linux_CentOS6-x86_64-gcc4.4.6/include/HDDM/hddm_s.hpp:428
0000017 0x00000000005bce25 in hddm_s::istream::sequencer (this=0x7fffec000d10, object=...)
   at /home/fs7/tboettch/HallD_Home/install/sim-recon/sim-recon-commissioning/Linux_CentOS6-x86_64-gcc4.4.6/include/HDDM/hddm_s.hpp:4961
0000018 0x00000000005bccaf in hddm_s::istream::operator>> (this=0x7fffec000d10, object=...)
   at /home/fs7/tboettch/HallD_Home/install/sim-recon/sim-recon-commissioning/Linux_CentOS6-x86_64-gcc4.4.6/include/HDDM/hddm_s.hpp:4951
0000019 0x00000000005bcec7 in hddm_s::istream::sequencer (this=0x7fffec000d10, object=<value optimized out>)
   at /home/fs7/tboettch/HallD_Home/install/sim-recon/sim-recon-commissioning/Linux_CentOS6-x86_64-gcc4.4.6/include/HDDM/hddm_s.hpp:4973
0000020 0x00000000005bccaf in hddm_s::istream::operator>> (this=0x7fffec000d10, object=...)
   at /home/fs7/tboettch/HallD_Home/install/sim-recon/sim-recon-commissioning/Linux_CentOS6-x86_64-gcc4.4.6/include/HDDM/hddm_s.hpp:4951
0000021 0x0000000000bca557 in hddm_s::istream::operator>> (this=0x7fffec000d10, record=...)
   at .Linux_CentOS6-x86_64-gcc4.4.6/libraries/HDDM/hddm_s++.cpp:217
0000022 0x0000000000b9cdc8 in DEventSourceHDDM::GetEvent (this=0x7fffec000fa0, event=...)
   at libraries/HDDM/DEventSourceHDDM.cc:109
0000023 0x0000000000c2f992 in jana::JEventSource::GetEvent (this=0x7fffec000fa0, event=...) at src/JANA/JEventSource.cc:54
0000024 0x0000000000bf174b in jana::JApplication::ReadEvent (this=0x7fffffffce20, event=...)
   at src/JANA/JApplication.cc:655
0000025 0x0000000000bf15c1 in jana::JApplication::EventBufferThread (this=0x7fffffffce20) at src/JANA/JApplication.cc:600
0000026 0x0000000000bf12af in LaunchEventBufferThread (arg=0x7fffffffce20) at src/JANA/JApplication.cc:552
0000027 0x0000003d99807851 in start_thread () from /lib64/libpthread.so.0
0000028 0x0000003d98ce811d in clone () from /lib64/libc.so.6_______________________________________________
Halld-offline mailing list
Halld-offline@jlab.org
https://mailman.jlab.org/mailman/listinfo/halld-offline [^]
Notes
(0000619)
davidl   
2014-12-03 15:44   
I was able to find one issue that looked to be causing the initial crashes Tom was seeing. It was in the smear_bcal.cc file and a fix was checked in to the commissioning branch as revision .

I was still unable to cleanly run mcsmear as it started crashing in a different place. This seems to be deep in the xdr reader code where it is getting a length that is equivalent to all bits set to 1 (i.e. very large number). This occurred around event 93 when I ran the following:

> mcsmear hdgeant_POmega.hddm

The backtrace is copied below from running on Linux. I also observed crashes at the same place when running on Mac OS X. It is worth noting that I also observe a problem using hddm-xml. It prints out a bunch of garbage for the "vals" field of bcalSiPMSpectrum structure in the event. This may indicate a corrupted output from hdgeant for the BCAL SiPM data.

----------------------------------------------------------------------------
Program received signal SIGSEGV, Segmentation fault.
[Switching to Thread 0x7fa4424b7700 (LWP 25944)]
0x0000000000d21ede in sgetn (this=0x7fa43c0218c0, s="") at /usr/lib/gcc/x86_64-redhat-linux/4.4.6/../../../../include/c++/4.4.6/streambuf:335
335 { return this->xsgetn(__s, __n); }
(gdb) bt
#0 0x0000000000d21ede in sgetn (this=0x7fa43c0218c0, s="") at /usr/lib/gcc/x86_64-redhat-linux/4.4.6/../../../../include/c++/4.4.6/streambuf:335
0000001 xstream::xdr::istream::operator>> (this=0x7fa43c0218c0, s="") at external/xstream/src/xdr.cpp:133
0000002 0x00000000005b7e4d in hddm_s::BcalSiPMSpectrum::streamer (this=0x7fa43c1ef1f0, istr=...)
    at /w/halld-scifs1a/home/davidl/builds/sim-recon-commissioning/Linux_CentOS6-x86_64-clang3.2/include/HDDM/hddm_s.hpp:16529
0000003 0x00000000005b74a5 in hddm_s::istream::sequencer (this=0x7fa43c000d60, object=...)
    at /w/halld-scifs1a/home/davidl/builds/sim-recon-commissioning/Linux_CentOS6-x86_64-clang3.2/include/HDDM/hddm_s.hpp:4961
0000004 0x00000000005cb651 in hddm_s::HDDM_ElementList<hddm_s::BcalSiPMSpectrum>::streamer (this=0x7fa43c5362a8, istr=...)
    at /w/halld-scifs1a/home/davidl/builds/sim-recon-commissioning/Linux_CentOS6-x86_64-clang3.2/include/HDDM/hddm_s.hpp:428
0000005 0x00000000005b74a5 in hddm_s::istream::sequencer (this=0x7fa43c000d60, object=...)
    at /w/halld-scifs1a/home/davidl/builds/sim-recon-commissioning/Linux_CentOS6-x86_64-clang3.2/include/HDDM/hddm_s.hpp:4961
0000006 0x00000000005b774d in hddm_s::istream::operator>> (this=0x7fa43c000d60, object=...)
    at /w/halld-scifs1a/home/davidl/builds/sim-recon-commissioning/Linux_CentOS6-x86_64-clang3.2/include/HDDM/hddm_s.hpp:4951
0000007 0x00000000005b7547 in hddm_s::istream::sequencer (this=0x7fa43c000d60, object=<value optimized out>)
    at /w/halld-scifs1a/home/davidl/builds/sim-recon-commissioning/Linux_CentOS6-x86_64-clang3.2/include/HDDM/hddm_s.hpp:4973
0000008 0x00000000005d11e7 in hddm_s::HDDM_ElementList<hddm_s::BcalCell>::streamer (this=<value optimized out>, istr=<value optimized out>)
    at /w/halld-scifs1a/home/davidl/builds/sim-recon-commissioning/Linux_CentOS6-x86_64-clang3.2/include/HDDM/hddm_s.hpp:428
0000009 0x00000000005b74a5 in hddm_s::istream::sequencer (this=0x7fa43c000d60, object=...)
    at /w/halld-scifs1a/home/davidl/builds/sim-recon-commissioning/Linux_CentOS6-x86_64-clang3.2/include/HDDM/hddm_s.hpp:4961
0000010 0x00000000005b774d in hddm_s::istream::operator>> (this=0x7fa43c000d60, object=...)
    at /w/halld-scifs1a/home/davidl/builds/sim-recon-commissioning/Linux_CentOS6-x86_64-clang3.2/include/HDDM/hddm_s.hpp:4951
0000011 0x00000000005b7547 in hddm_s::istream::sequencer (this=0x7fa43c000d60, object=<value optimized out>)
    at /w/halld-scifs1a/home/davidl/builds/sim-recon-commissioning/Linux_CentOS6-x86_64-clang3.2/include/HDDM/hddm_s.hpp:4973
#12 0x00000000005b774d in hddm_s::istream::operator>> (this=0x7fa43c000d60, object=...)
    at /w/halld-scifs1a/home/davidl/builds/sim-recon-commissioning/Linux_CentOS6-x86_64-clang3.2/include/HDDM/hddm_s.hpp:4951
0000013 0x00000000005b7547 in hddm_s::istream::sequencer (this=0x7fa43c000d60, object=<value optimized out>)
    at /w/halld-scifs1a/home/davidl/builds/sim-recon-commissioning/Linux_CentOS6-x86_64-clang3.2/include/HDDM/hddm_s.hpp:4973
0000014 0x00000000005b774d in hddm_s::istream::operator>> (this=0x7fa43c000d60, object=...)
    at /w/halld-scifs1a/home/davidl/builds/sim-recon-commissioning/Linux_CentOS6-x86_64-clang3.2/include/HDDM/hddm_s.hpp:4951
0000015 0x00000000005b7547 in hddm_s::istream::sequencer (this=0x7fa43c000d60, object=<value optimized out>)
    at /w/halld-scifs1a/home/davidl/builds/sim-recon-commissioning/Linux_CentOS6-x86_64-clang3.2/include/HDDM/hddm_s.hpp:4973
0000016 0x00000000005c39bf in hddm_s::HDDM_ElementList<hddm_s::PhysicsEvent>::streamer (this=<value optimized out>, istr=...)
    at /w/halld-scifs1a/home/davidl/builds/sim-recon-commissioning/Linux_CentOS6-x86_64-clang3.2/include/HDDM/hddm_s.hpp:428
0000017 0x00000000005b74a5 in hddm_s::istream::sequencer (this=0x7fa43c000d60, object=...)
    at /w/halld-scifs1a/home/davidl/builds/sim-recon-commissioning/Linux_CentOS6-x86_64-clang3.2/include/HDDM/hddm_s.hpp:4961
0000018 0x00000000005b774d in hddm_s::istream::operator>> (this=0x7fa43c000d60, object=...)
    at /w/halld-scifs1a/home/davidl/builds/sim-recon-commissioning/Linux_CentOS6-x86_64-clang3.2/include/HDDM/hddm_s.hpp:4951
0000019 0x00000000005b7547 in hddm_s::istream::sequencer (this=0x7fa43c000d60, object=<value optimized out>)
    at /w/halld-scifs1a/home/davidl/builds/sim-recon-commissioning/Linux_CentOS6-x86_64-clang3.2/include/HDDM/hddm_s.hpp:4973
0000020 0x00000000005b774d in hddm_s::istream::operator>> (this=0x7fa43c000d60, object=...)
    at /w/halld-scifs1a/home/davidl/builds/sim-recon-commissioning/Linux_CentOS6-x86_64-clang3.2/include/HDDM/hddm_s.hpp:4951
0000021 0x0000000000cf3a35 in hddm_s::istream::operator>> (this=0x7fa43c000d60, record=...) at .Linux_CentOS6-x86_64-clang3.2/libraries/HDDM/hddm_s++.cpp:217
0000022 0x0000000000ca89c6 in DEventSourceHDDM::GetEvent (this=0x7fa43c000f70, event=...) at libraries/HDDM/DEventSourceHDDM.cc:109
0000023 0x0000000000d65576 in jana::JEventSource::GetEvent (this=0x7fa43c000f70, event=...) at src/JANA/JEventSource.cc:54
0000024 0x0000000000d2711b in jana::JApplication::ReadEvent (this=0x7fffe24bc6d0, event=...) at src/JANA/JApplication.cc:655
0000025 0x0000000000d26f91 in jana::JApplication::EventBufferThread (this=0x7fffe24bc6d0) at src/JANA/JApplication.cc:600
0000026 0x0000000000d26c7f in LaunchEventBufferThread (arg=0x7fffe24bc6d0) at src/JANA/JApplication.cc:552
0000027 0x00000033466077f1 in start_thread () from /lib64/libpthread.so.0
0000028 0x0000003345ee5ccd in clone () from /lib64/libc.so.6
(gdb)
(0000647)
davidl   
2015-06-10 09:42   
This looks like it may have been fixed by Sean D. with revision 18034 on April 15, 2015. I've not heard recent reports of mcsmear crashes.





View Issue Details
427 [Hall D Offline] General minor N/A 2014-12-03 08:08 2015-06-10 09:36
davidl  
davidl  
normal  
resolved  
fixed  
none    
none  
   
Extract event number from EVIO for use in JEvent
Currently, the DAQ plugin sets the event number using a counter. This should be changed to use the event number stored in the event itself. The JEventSource_EVIO::FindRunNumber method will need to be modified to return the event number as well. Some care will need to be taken to calculate the event number for stacked events.
Notes
(0000646)
davidl   
2015-06-10 09:36   
This was implemented in revision 17107 on Jan 26, 2015





View Issue Details
408 [Hall D Offline] General minor have not tried 2014-04-18 10:42 2015-06-10 09:34
marki  
davidl  
normal  
resolved  
fixed  
none    
none  
   
look into upgrading wiki version
from 4/16/14 meeting
Notes
(0000645)
davidl   
2015-06-10 09:34   
This was done by Marty months ago.





View Issue Details
416 [JANA] Feature Request minor N/A 2014-09-29 07:24 2015-06-09 23:20
davidl  
davidl  
normal  
resolved  
fixed  
none    
none  
   
Dump factory call stack if factory not found
If possible, the factory call stack should be printed if a request for an unknown factory is encountered. This will make it easier to debug where the request is coming from.
Notes
(0000644)
davidl   
2015-06-09 23:20   
This was implemented in revision 2339 and will appear in JANA 0.7.3. The actual change needed was to throw an exception when the error was detected rather than simply quit the program.





View Issue Details
447 [JANA] Feature Request minor N/A 2015-01-26 08:12 2015-06-09 16:52
davidl  
davidl  
normal  
resolved  
fixed  
none    
none  
   
JEvent:event_number should be 64 bit
Currently, the type of the event_number member of JEvent is "int" which is a signed 32 bit value. This is insufficient for runs with over 2billion events. This can happen if we realize a 200kHz trigger rate for a run lasting 3 hours. Note that even though 200kHz won't be written to the file, we may preserve the original event number going into the L3 trigger so the event numbers can exceed the 2billion limit. Also, the EVIO format stores the event number as a 64bit value in the Trigger bank so we need a 64 bit value to maintain compatibility.
Notes
(0000643)
davidl   
2015-06-09 16:52   
This was implemented in revision 2338 and will appear in JANA 0.7.3. The change leaves the "int" versions of JEventProcessor::brun and JEventProcessor::evnt that are actually called by the new versions that use uint32_t and uint64_t respectively. This should allow sim-recon to work with JANA 0.7.3 without having to update every JEventProcessor and JFactory in the code right away.





View Issue Details
426 [JANA] Feature Request minor N/A 2014-12-03 06:18 2015-06-09 16:15
davidl  
davidl  
normal  
resolved  
fixed  
none    
none  
   
Add event barrier
There are situations where special events are encountered in the data stream that may affect how the events following it are analyzed. These include EPICS events and sync events. In order to ensure that all subsequent events see the effects of the special event, but no previous ones still being processed do, a barrier mechanism is needed.

The barrier should be callable from the JEventSource::GetEvent() method and should let the event buffer drain completely before continuing.
Notes
(0000642)
davidl   
2015-06-09 16:15   
This was implemented in revision 2337 and will appear in JANA 0.7.3. If an event source calls:

event.SetSequential()

in its GetEvent method than the event is flagged to be run sequentially. A unit test called "event_barrier" was created to test the mechanism and it appears to work.





View Issue Details
453 [JANA] Feature Request minor have not tried 2015-02-25 08:11 2015-06-09 10:51
davidl  
davidl  
normal  
resolved  
fixed  
none    
none  
   
Config. Parameter to override run number
It may be convenient to have a configuration parameter that can be used to override whatever run number is provided by the event source. This would need to be applied in the JApplication class when the event is read in by the event reader thread calling GetEvent().
Notes
(0000641)
davidl   
2015-06-09 10:51   
A config. parameter RUNNUMBER was implemented.

This was implemented in revision 2336 and will appear in JANA 0.7.3





View Issue Details
455 [JANA] Bug major always 2015-02-28 19:33 2015-06-05 16:23
sdobbs  
davidl  
high  
resolved  
fixed  
none    
none  
   
Wrong ordering when loading CCDB tables with multiple columns
When loading tables with multiple columns in JANA, the columns are loaded ordered by their column labels alphabetically, instead of the order they are stored in the CCDB.

This problem came up when loading the CCDB table /BCAL/attenuation_parameters, which has three columns, named "L1", "L2", and "L0", respectively. When loading the table in DBCALPoint_factory, the values were loaded in the order "L0", "L1", "L2".
This problem is illustrated in the plugin in the following location:

  /w/halld-scifs1a/home/sdobbs/plugins/CCDBTest

compared to the output of

  ccdb cat /BCAL/attenuation_parameters

Sample outputs are attached.
Dmitry pointed out that the problem lies in this function:

//-------------
template<class T> bool JCalibration::Get(string namepath, vector< vector<T> > &vals, int event_number)
{
    vector< map<string, string> >svals;
    bool res = GetCalib(namepath, svals, event_number);
  
    // copying them into the vals map.
    vals.clear();
    for(unsigned int i=0; i<svals.size(); i++){
        map<string,string>::const_iterator iter;
  ...
    }
    
    return res;
}

Since this function gets the calibration values as a map instead of a vector<vector<>>, and maps in C++ are ordered by key name, this is where this mis-ordering comes from.
bug_output.txt (32,538) 2015-02-28 19:33
https://halldweb.jlab.org/mantisbt/file_download.php?file_id=4&type=bug
Notes
(0000629)
davidl   
2015-03-02 10:17   
This is a feature designed into the API for JANA. The intent is to minimize the number of virtual methods that must be implemented in a class inheriting from JCalibration. The design was to allow one to access a table that had columns defined by name using:

   Get(string name path, vector< map<string, T> > &vals, int event_number)

And columns defined by position using:

   Get(string name path, vector< vector<T> > &vals, int event_number)


The idea was that if you had a table where the columns did not have names and were only defined by position, then the underlying DB was responsible for maintaining the ordering and one would request the table using the "vector" form of Get(). If had a table with named columns, then you should be using the "map" form of Get() which gives you the values indexed by column name. The only way to fix this so that one may use the "vector" form with named columns and still preserve order is to add another pair of GetCalib virtual methods to JCalibration and have CCDB implement those.
(0000640)
davidl   
2015-06-05 16:23   
This was implemented in revision 2334 and will appear in JANA 0.7.3. It also required changes to DCalibrationCCDB that exists in the DANA library in sim-recon. Those changes have already been checked in and checked that they still compile with JANA 0.7.2 this shouldn't require a synchronous upgrade between JANA and sim-recon.





View Issue Details
457 [JANA] Bug minor always 2015-03-17 13:15 2015-06-05 11:06
kmoriya  
davidl  
low  
resolved  
fixed  
none    
none  
   
jana fails to compile if BMS_OSNAME is not previously set
Within the SConstruct file in jana, the variable osname is
being attempted to be set by SMBS/osrelease.pl if it has not
been set already. This fails since the file is only in the BMS directory.
Notes
(0000639)
davidl   
2015-06-05 11:06   
Moved the osrelease.pl file from BMS to SMBS directory and removed the BMS directory altogether. This was committed in revision 2333 and will appear in JANA 0.7.3.





View Issue Details
452 [JANA] Feature Request minor N/A 2015-02-12 13:49 2015-06-05 10:55
davidl  
davidl  
normal  
resolved  
fixed  
none    
none  
   
Add special handling of -PAUTOACTIVATE=ALL
When running hd_ana with the janaroot plugin on an hddm file, I wanted to make trees for all available objects. This could be done if JANA recognized the value "ALL" for the AUTOACTIVATE config parameter as meaning all factories. One could also implement another like "SOURCE" that would specify only objects that come from the event source (though, come to think of it, that may not be possible).
Notes
(0000638)
davidl   
2015-06-05 10:55   
This was added while addressing 420. It was originally implemented to handle only "all" but with revision 2332 is made to also accept "ALL". This will appear in JANA 0.7.3





View Issue Details
448 [JANA] Feature Request minor N/A 2015-01-26 08:18 2015-06-05 10:50
davidl  
davidl  
normal  
resolved  
fixed  
none    
none  
   
Easier way to print configuration parameters.
The current system for discovering configuration parameters from an executable requires the user to run the executable for at least one event with the --dumpconfig option and then examine the resulting jana.config file. A simpler option is needed to try and quickly get at this list. Unfortunately, generating the complete list requires calling init and brun routines which perform all kinds of expensive initializations. Nonetheless, the situation can still be improved upon.
Notes
(0000637)
davidl   
2015-06-05 10:50   
duplicate of issue 420





View Issue Details
420 [JANA] Feature Request minor N/A 2014-11-10 10:14 2015-06-05 10:50
davidl  
davidl  
normal  
resolved  
fixed  
none    
none  
   
add --listconfig more useful
The --dumpconfig option is somewhat limited in usefulness to find out what parameters exist since it needs to have an event source to run and most factories need to be activated in order to register their parameters. A "--listconfig" option could be used to activate all factories and dump the resulting list of all parameters to the screen in a way similar to --dumpconfig.

One could imagine two default behaviors:

1.) If no JEventProcessors are registered and no AUTOACTIVATE factories exist, then try activating all factories in order to get a complete list

2.) If at least one JEventProcessor exists or an AUTOACTIVATE factory is specified, then only print parameters registered via those.

This has the disadvantage of suppressing parameters from being registered/printed if something is *added* to the command line.

Perhaps one wants both a "--listconfig" and a "--listallconfig" option so that both behaviors are available and the user selects which they want(??)
Notes
(0000636)
davidl   
2015-06-05 10:49   
This was implemented in revision 2331 and will appear in JANA 0.7.3.

It was not implemented exactly as described. Only a --listconfig options exists and it always automatically activates all factories. It also replaces the source source list with a dummy one and forces the use of the JEventSource_NULL source type with 1 thread and 1 event. This allows one to type the following to get a complete list:

hd_ana --listconfig





View Issue Details
454 [JANA] Feature Request minor N/A 2015-02-28 19:03 2015-06-04 14:34
sdobbs  
davidl  
normal  
resolved  
fixed  
none    
none  
   
Access to CCDB variation information inside JANA
It would be useful if to be able to access the CCDB variation name in the Calibration object.

For example, if JANA_CALIB_CONTEXT="variation=mc", there could be a function that returned "mc". If not explicitly set, the reported variation would be "default".
Notes
(0000635)
davidl   
2015-06-04 14:34   
This was committed in revision 2328 and will appear in JANA 0.7.3





View Issue Details
415 [JANA] Feature Request minor N/A 2014-08-23 22:44 2015-06-04 14:34
davidl  
davidl  
normal  
resolved  
fixed  
none    
none  
   
Add config. param. to allow users to force a specific JEventSource be used.
When dealing with HDDM files, an issue came up when we tried using named pipes. Specifically, the CheckOpenable methods used for the "s" type (=simulation) and "r" type (=REST) sources tested them by trying to open the file. If a named pipe was being used, then the header was pulled off of the stream but the test open causing the second open to fail. To get around this, Richard implemented an algorithm, that looks for key phases in the filename itself, thereby avoiding the test open. The problem with this method is that it can be confused by a poorly named file.

In order to provide a general mechanism for users to side step issues such as this, a configuration parameter should be introduced that allows the user to specify exactly which JEventSource should be instantiated to handle the source. Perhaps something like "EVENT_SOURCE_TYPE" (?)
Notes
(0000634)
davidl   
2015-06-04 14:34   
This was committed in revision 2329 and will appear in JANA 0.7.3





View Issue Details
458 [JANA] Feature Request minor have not tried 2015-05-29 09:16 2015-06-01 16:01
davidl  
davidl  
normal  
resolved  
fixed  
none    
none  
   
Access to Configuration Parameter list programmatically
This came up in an e-mail exchange with Naomi on May 29th, 2015.

She was writing a plugin to create a ROOT tree with all configuration parameters. This would allow them to be stored within the ROOT file itself along with the histograms and trees produced using those parameters. JParameterManager needs to provide a method that returns a list of all parameters.
Notes
(0000633)
davidl   
2015-06-01 16:01   
Somehow I missed that the GetParameters() method of JParameterManager has been in existence since 2007. Oops!





View Issue Details
417 [JANA] Feature Request minor N/A 2014-09-29 07:30 2015-06-01 15:54
davidl  
davidl  
normal  
resolved  
fixed  
none    
none  
   
Add GetRecursive method to JObject
It would be useful to be able to ask JANA to not only search associated objects, by the associated objects of each associated object when looking for a specific type. e.g. asking a DBCALHit for the Df250PulseIntegral it is associated with.

Care would need to be taken not to run into an infinite loop. An optional parameter would probably also be needed to allow the caller to specify a maximum depth.
Notes
(0000632)
davidl   
2015-06-01 15:54   
This was added in revision 2322 and will appear in JANA 0.7.3. This was actually added to the already existing Get() method and an optional parameter added to that call to specify the maximum depth of associations to check. This is set to a very large number by default so that for practical purposes, all levels are checked. To get the original behavior of only checking direct descendants,, users must specify max_depth=1.





View Issue Details
421 [JANA] Feature Request minor N/A 2014-11-10 10:19 2015-06-01 15:49
davidl  
davidl  
normal  
resolved  
fixed  
none    
none  
   
Make JANA_CALIB_URL a configuration parameter
Paul Mattione suggested making a JANA_CALIB_URL configuration parameter that could override the environment variable of the same name. This would make it a little easier to organize configurations since that info could then be specified in the jana configuration file
Notes
(0000631)
davidl   
2015-06-01 15:49   
Committed in revision 2322 and will appear in JANA 0.7.3





View Issue Details
435 [Hall D Offline] General minor N/A 2014-12-11 19:59 2015-04-27 11:08
davidl  
davidl  
normal  
resolved  
fixed  
none    
none  
   
Need 800A solenoid map
From Kei:

Hi David,

I was running the offline monitoring on recent runs,
and noticed there are some B-field settings we don't
have maps for yet.

It probably doesn't make sense to create maps for each
setting we took data at (there's single runs with 400 A,
950 A), but we had a bunch of 800 A runs on Dec 4
(runs 1680 - 1696).

I don't know how much effort it takes to create these
maps, but could you create the 800 A version when you
have time?

Thanks,
    Kei
Notes
(0000630)
davidl   
2015-04-27 11:08   
Done. Map is Magnets/Solenoid/solenoid_800A_poisson_20150427





View Issue Details
449 [Hall D Offline] General minor always 2015-01-26 08:21 2015-01-26 08:21
davidl  
 
normal  
new  
open  
none    
none  
   
janadot graph not correct for EVIO source
This is a simple fix that was disabled due to needing a feature added to a later version of JANA. (maybe 0.7.2?) This version has been out for a while now so it should be fair to re-activate the fix so the call graph draws correctly.

The HTML map in the Doxygen documentation should be updated with a new graph once this is fixed.
There are no notes attached to this issue.





View Issue Details
446 [Hall D Offline] General minor sometimes 2015-01-20 16:57 2015-01-20 16:57
kmoriya  
 
normal  
new  
open  
none    
none  
   
Unknown module type error
When processing evio file, the following type of message appears:

JANA ERROR>>Unknown module type (9) iptr=0x0x7f26a0716680
JANA ERROR>>...skipping to 0x0x7ffc33b4cc5c (discarding 3 words)

The module type is between 6 and 15.

Examples of files showing this behavior and no other errors are
hd_rawdata_002239_000.evio
hd_rawdata_002030_002.evio
There are no notes attached to this issue.





View Issue Details
445 [Hall D Offline] General minor have not tried 2015-01-15 16:41 2015-01-15 16:47
davidl  
davidl  
normal  
resolved  
fixed  
none    
none  
   
SBMS does not automatically recompile when rest.xml is changed
From Paul M. in e-mail on 1/15/2015

---------------------------------------
Hey David, I noticed a problem with SBMS: Once I’ve compiled everything, if I then edit the rest.xml file, it doesn’t automatically recompile the hddm header files (or if it did, it didn’t install them).

- Paul
---------------------------------------
Notes
(0000628)
davidl   
2015-01-15 16:47   
seems like a false alarm. It is working for Paul now and I checked it as well.





View Issue Details
444 [Hall D Offline] General minor N/A 2015-01-05 14:05 2015-01-08 14:27
davidl  
davidl  
normal  
resolved  
fixed  
none    
none  
   
Move DAQ and TTab plugins into sim-recon library or libraries
The DAQ and TTab plugins are now in regular use and should be part of the standard tools so that they don't have to be specified on the command line every time a program is run. A major constraint is that the DAQ plugin should still exist so that it can be used with generic JANA programs (e.g. jana, janadump, ...). The TTab plugin, by its nature, contains detector-specific classes and so can't really be used with generic JANA tools. Therefore, only DAQ needs a plugin representation.
Notes
(0000627)
davidl   
2015-01-08 14:27   
Done. Committed Jan. 7th to commissioning branch.





View Issue Details
423 [Hall D Online] General major always 2014-12-02 16:12 2015-01-05 13:58
davidl  
davidl  
high  
resolved  
fixed  
none    
none  
   
No
0
0
Corrupted EVIO files?
Some fraction of data files have an issue that part way in, the EVIO library gets confused. Initial debugging looks like it is interpreting a word from the middle of a data block as a length word for the next EVIO event. It's unclear if this is due to corruption of the file or a bug in the library reading it. Regardless, the DAQ plugin should try and work around this error.

Richard Jones reported getting the error at around event 1000 of hd_rawdata_001411_000.evio.
Notes
(0000626)
davidl   
2015-01-05 13:58   
This was addressed with the introduction of the HDEVIO parser into the DAQ plugin source on 12-10/12-11. It was made the default parser on 12-16-2014.





View Issue Details
443 [Hall D Offline] General minor always 2015-01-05 13:40 2015-01-05 13:41
davidl  
davidl  
normal  
resolved  
won't fix  
none    
none  
   
Mismatch of number of events passed to MergeObjLists Exception
Paul M. and Mark D. both reported seeing this issue for runs 1518 and 2205 respectively. The symptom is an exception message such as the following. IT was reported for runs 1518, 2200, and 2205, though they are probably the tip of the iceberg.

Mon Dec 22 18:56:12 2014 # thr=139666653529856 # --- EVIO ---: Mismatch of number of events passed to MergeObjLists. Throwing exception.
Mon Dec 22 18:56:12 2014 # thr=139666653529856 # --- EVIO ---: Nevents1=1 Nevents2=2
JANA ERROR>>Exception thrown from ParseEVIOEvent!
JANA ERROR>>
JANA ERROR>>?JException: code = 0 text = Number of events in JEventSource_EVIO::MergeObjLists do not match!
JANA ERROR>>
JANA ERROR>>Stack trace:
JANA ERROR>>
JANA ERROR>> jana::JException::getStackTrace(bool, unsigned long)
JANA ERROR>> jana::JException::JException(std::string const&)
JANA ERROR>> JEventSource_EVIO::MergeObjLists(std::list<JEventSource_EVIO::ObjList*, std::allocator<JEventSource_EVIO::ObjList*> >&, std::list<JEventSource_EVIO::ObjList*,
Notes
(0000625)
davidl   
2015-01-05 13:41   
Below is a message sent to the halld-offline mailing list regarding this on Jan. 5, 2015:

Hi Offliners,

  I’ve received a couple of reports about errors containing lines similar to the ones below when processing
commissioning data. This problem seems to be hardware related. A single fADC125 module for the CDC
occasionally had a flaky bit 19 in the itrigger value it reported. This causes the code to treat it as a separate
event, thus the mismatch in “events" coming from this module compared to all others. As such, the error message
is actually valid and correct. This issue can be ameliorated by setting the configuration parameter EVIO:ENABLE_DISENTANGLING
to “0”. e.g.

hd_root -PEVIO:ENABLE_DISENTANGLING=0 …

This only works for files taken with CODA block level=1. For all useful commissioning data this was the case.

We can discuss how best to handle these types of errors since automatically suppressing them is not really
good. However, specifying the above often can be a bit annoying too. Suggestions are welcome.





View Issue Details
440 [Hall D Online] DAQ major sometimes 2014-12-16 16:49 2014-12-30 15:41
davidl  
 
normal  
new  
open  
none    
none  
   
No
0
0
Files on tape not matching ones on RAID
After discussing with David what we should do to verify that
data is being correctly written to tape, I decided to
use the md5checksum. This information is created and according
to David, checked when files are written to tape. So it seemed
like a duplicated effort, but I went ahead anyway.

The numbers I am comparing are
1. the line that starts with "md5=" within each file for
/mss/halld/RunPeriod-2014-10/rawdata/Run002104/*.evio
2. running md5sum on each file on the evio files on the gluonraids

Running md5sum is rather cumbersome since you have to compute
this value based on every single byte in the file. This can
take between 1-3 min. for each file, so to save time I restricted
myself to files ending in _000.evio.

The result is, I find 5 pairs where the md5sum do not agree.
The runs and md5sums are below. The "<" files are from /mss,
the ">" files are from running on the gluonraid files.

< hd_rawdata_000990_000.evio c68868a8d28e04f9a9c3e1da69b16f08
> hd_rawdata_000990_000.evio 289c814000df3c24c9b21dd8b9db224b

< hd_rawdata_001291_000.evio 6f6bf352b2549a0eb5fc2b95b56fafe3
> hd_rawdata_001291_000.evio 46d356ff5033b0565bf0a2d361f72745

< hd_rawdata_001398_000.evio 4d940f220dba37209f3a2749fa8c6e40
> hd_rawdata_001398_000.evio d082eb4b27035e9034f0671011ede4f4

< hd_rawdata_001642_000.evio 5b3283c1a361e4a41ff16c690faa8849
> hd_rawdata_001642_000.evio 49035579053149f1d6879a98a7c0c21c

< hd_rawdata_002104_000.evio 69c7515b40359444b6c0afd82ba3908d
> hd_rawdata_002104_000.evio a75a12de0f8fa4b17cc8729dc1fd81b5

I ran over the 000 files between 940 and 2245, but note that
there are actually some runs where file 000 is missing for one
reason or the other (DAQ crash etc). So for roughly 1300 files,
I find 5 files (~0.4%) have wrong md5 checksums.

I checked the size of each file, and each pair seems to have different
values. The below shows file names, size shown in /mss, and size in
gluonraid. The gluonraid files are always bigger.

000990_000.evio 16132156448 16132463696
001291_000.evio 239303640 397829944
001398_000.evio 5645101228 5645842860
001642_000.evio 5707567564 5709949640
002104_000.evio 715201656 1014051452

The runs that show a difference don't seem to be clustered
at any specific value. Since there is now the potential that
there is something wrong in the copying process, we may
need to think of how to eliminate the bug.

    Kei
I checked the file sizes of all evio files in the gluonraid disks
and the /mss system.

Below are the 17 files that do not agree in size:

< hd_raw_000691_000.evio 797324740
> hd_raw_000691_000.evio 1734616380

> hd_raw_000692_000.evio 5411961244
< hd_raw_000692_000.evio 797095824

< hd_raw_000696_000.evio 818140800
> hd_raw_000696_000.evio 1964263484

< hd_raw_000697_000.evio 938227060
> hd_raw_000697_000.evio 949686472

< hd_raw_000700_000.evio 818631544
> hd_raw_000700_000.evio 4117328924

< hd_raw_000702_000.evio 820366980
> hd_raw_000702_000.evio 3363271288

< hd_raw_000707_000.evio 1292643512
> hd_raw_000707_000.evio 2594874824

< hd_raw_000710_000.evio 1290320092
> hd_raw_000710_000.evio 2498216628

< hd_raw_000718_000.evio 1291617672
> hd_raw_000718_000.evio 10043266080

< hd_raw_000731_000.evio 1101749368
> hd_raw_000731_000.evio 1312180216

< hd_raw_000803_000.evio 1232781032
> hd_raw_000803_000.evio 1565069964

< hd_rawdata_000990_000.evio 16132156448
> hd_rawdata_000990_000.evio 16132463696

< hd_rawdata_001291_000.evio 239303640
> hd_rawdata_001291_000.evio 397829944

< hd_rawdata_001398_000.evio 5645101228
> hd_rawdata_001398_000.evio 5645842860

< hd_rawdata_001642_000.evio 5707567564
> hd_rawdata_001642_000.evio 5709949640

< hd_rawdata_001656_036.evio 5229012312
> hd_rawdata_001656_036.evio 5229903160

< hd_rawdata_002104_000.evio 715201656
> hd_rawdata_002104_000.evio 1014051452
Notes
(0000624)
sdobbs   
2014-12-30 15:41   
(edited on: 2014-12-30 15:42)
I checked the md5 checksums of all the EVIO files on the RAID disk with those reported by the tape system in the stub files in /mss [Note that there were a handful of files my script didn't find on the RAID disk (about 20)]. Of the analyzed files, the files whose checksums did not match overlaps almost perfectly with Kei's list - there is one extra file which presumably was created after his analysis.

This is consistent with the hypothesis that the files are being copied to tape without error, and the mismatches we see are only due to files on tape being truncated.

                filename RAID checksum tape checksum

        hd_raw_000691_000.evio 39f677ec9417a8b32e07773f780e2b4e 9f2a3f3289e77dc1240c1794d6b9b262
        hd_raw_000692_000.evio 012b6a53f7d0ef1854884a92bd1e4048 17e1afb6359c7582ed7c680224d42522
        hd_raw_000696_000.evio 4ad2e8b77ef8eb5c9746bddc0e7f76af 03e005874f048917d5f71235ccfd6ada
        hd_raw_000697_000.evio 7bc86001ca2b561c29640f9024626e6e d4b548fb4db62096d0114f34594f3641
        hd_raw_000700_000.evio e58acfc4cea2e4ab8d8a7a43c77af08a 798db9c2a0e359384062123f41166820
        hd_raw_000702_000.evio 3c71b0cd9642165368a11ab24cfd81f4 90e84753f2558ee1c871c5c849ce42de
        hd_raw_000707_000.evio 885c811f69c957b9fd31bf8d9619f928 e555119d6731a1bef043fc9197186d72
        hd_raw_000710_000.evio 23be1029c2194bd4f4aec9f4b851ba98 26666c83f2ef6f4c2926571475d003c3
        hd_raw_000718_000.evio b67fb9900b75de21bb6d7d7395f69193 084007a8ea31915c7f421eaac97a351c
        hd_raw_000731_000.evio b4ad2166a83f2ce25e383614f09d0c68 024772c69d13a21d708f766eb4d032f8
        hd_raw_000803_000.evio 44bbb75ed7f0240e2e8b8421eff3a85c f4da62d69ddfe8ec0f1fd1852150fdc5
    hd_rawdata_000990_000.evio 289c814000df3c24c9b21dd8b9db224b c68868a8d28e04f9a9c3e1da69b16f08
    hd_rawdata_001291_000.evio 46d356ff5033b0565bf0a2d361f72745 6f6bf352b2549a0eb5fc2b95b56fafe3
    hd_rawdata_001398_000.evio d082eb4b27035e9034f0671011ede4f4 4d940f220dba37209f3a2749fa8c6e40
    hd_rawdata_001642_000.evio 49035579053149f1d6879a98a7c0c21c 5b3283c1a361e4a41ff16c690faa8849
    hd_rawdata_001656_036.evio c17fc4b1c1a81d0b66bf69a54295a5fd 0450214ed7d71c09be5dfeec6e11a595
    hd_rawdata_002104_000.evio a75a12de0f8fa4b17cc8729dc1fd81b5 69c7515b40359444b6c0afd82ba3908d
    hd_rawdata_002361_000.evio 744558a07d81abcbe81523befd74e8bb 14189d779a6c62c4cdb16b0aeb9c1b96






View Issue Details
442 [Hall D Online] Monitoring major always 2014-12-20 23:11 2014-12-20 23:11
davidl  
 
normal  
new  
open  
none    
none  
   
No
0
0
Slow event processing rate for monitor programs
The monitoring system processes events at a much slower rate than the CPU is capable of. This may be helped if the monitoring is done using a secondary ET system. This will also help reduce the number of et clients connected to the DAQ-critical ET system used but the ER.
There are no notes attached to this issue.





View Issue Details
441 [Hall D Offline] General major sometimes 2014-12-18 10:58 2014-12-18 10:58
davidl  
 
normal  
new  
open  
none    
none  
   
CCDB crash
Reported by Mike Staib on Dec. 18, 2014 in the following e-mail

Hi,

I figured I would report this crash since it might be a bug that is hanging around. I couldn’t reproduce it, so it happens spuriously. Since I caught the backtraces, some of this might be interesting. It might be a while before this one is caught out in the wild again. Not a high priority to fix:

Every thread was in pthread_mutex_lock() except this one:

Thread 5 (Thread 0x7f4b89365700 (LWP 9890)):
#0 0x00000032b68ac65d in waitpid () from /lib64/libc.so.6
0000001 0x00000032b683e609 in do_system () from /lib64/libc.so.6
0000002 0x00000032b683e940 in system () from /lib64/libc.so.6
0000003 0x00007f4ba20a6298 in TUnixSystem::StackTrace() () from /home/gluex/root/v5-34-14_rhel6//lib/libCore.so
0000004 0x00007f4ba20a37ef in TUnixSystem::DispatchSignals(ESignals) () from /home/gluex/root/v5-34-14_rhel6//lib/libCore.so
0000005 <signal handler called>
0000006 0x00000032bdc6a08a in std::_Rb_tree_insert_and_rebalance(bool, std::_Rb_tree_node_base*, std::_Rb_tree_node_base*, std::_Rb_tree_node_base&) () from /usr/lib64/libstdc++.so.6
0000007 0x00007f4b9d4f298e in std::_Rb_tree<unsigned long, std::pair<unsigned long const, ccdb::StoredObject*>, std::_Select1st<std::pair<unsigned long const, ccdb::StoredObject*> >, std::less<unsigned long>, std::allocator<std::pair<unsigned long const, ccdb::StoredObject*> > >::_M_insert (this=0x7f4b7411e2d8, __x=<value optimized out>, __p=0x7f4b1db02860, __v=...) at /usr/local/lib/gcc/x86_64-unknown-linux-gnu/4.2.3/../../../../include/c++/4.2.3/bits/stl_tree.h:840
0000008 0x00007f4b9d4f2c10 in std::_Rb_tree<unsigned long, std::pair<unsigned long const, ccdb::StoredObject*>, std::_Select1st<std::pair<unsigned long const, ccdb::StoredObject*> >, std::less<unsigned long>, std::allocator<std::pair<unsigned long const, ccdb::StoredObject*> > >::_M_insert_unique (this=0x0, __position=<value optimized out>, __v=...) at /usr/local/lib/gcc/x86_64-unknown-linux-gnu/4.2.3/../../../../include/c++/4.2.3/bits/stl_tree.h:1006
0000009 0x00007f4b9d4f25dc in insert (this=<value optimized out>, object=0x7f4b1dc12188) at /usr/local/lib/gcc/x86_64-unknown-linux-gnu/4.2.3/../../../../include/c++/4.2.3/bits/stl_map.h:427
0000010 operator[] (this=<value optimized out>, object=0x7f4b1dc12188) at /usr/local/lib/gcc/x86_64-unknown-linux-gnu/4.2.3/../../../../include/c++/4.2.3/bits/stl_map.h:350
0000011 ccdb::ObjectsOwner::BeOwner (this=<value optimized out>, object=0x7f4b1dc12188) at src/Library/Model/ObjectsOwner.cc:40
#12 0x00007f4b9d4f3172 in ccdb::Assignment::Assignment (this=0x7f4b1dc12150, owner=0x7f4b7411e2d0, provider=0x7f4b7411e2d0) at src/Library/Model/Assignment.cc:21
0000013 0x00007f4b9d502e20 in ccdb::SQLiteDataProvider::GetAssignmentShort (this=0x7f4b7411e2d0, run=1505, path="/FDC/package3/strip_gains", time=0, variationName=<value optimized out>, loadColumns=<value optimized out>) at src/Library/Providers/SQLiteDataProvider.cc:1181
0000014 0x00007f4b9d4e5f05 in ccdb::Calibration::GetAssignment (this=0x7f4b7411b6c0, namepath=<value optimized out>, loadColumns=true) at src/Library/Calibration.cc:607
0000015 0x00007f4b9d4e3497 in ccdb::Calibration::GetCalib (this=0x0, values=std::vector of length 0, capacity 0, namepath=Traceback (most recent call last):
   return self.val['_M_dataplus']['_M_p'].lazy_string (length = len)
0000016 0x0000000000bec9ab in jana::JCalibrationCCDB::GetCalib (this=0x7f4b7411b710, namepath="/FDC/package3/strip_gains", vsvals=std::vector of length 0, capacity 0, event_number=3) at src/JANA/JCalibrationCCDB.h:133
0000017 0x00000000009453c0 in jana::JCalibration::Get<double> (this=0x7f4b7411b710, namepath="/FDC/package3/strip_gains", vals=std::vector of length 0, capacity 0, event_number=3) at /home/mstaib/gluex/trunk/jana_0.7.2/Linux_RHEL6-x86_64-gcc4.4.7/include/JANA/JCalibration.h:314
0000018 0x0000000000945a7b in jana::JEventLoop::GetCalib<std::vector<double, std::allocator<double> > > (this=0x7f4b6c0008c0, namepath="/FDC/package3/strip_gains", vals=std::vector of length 0, capacity 0) at /home/mstaib/gluex/trunk/jana_0.7.2/Linux_RHEL6-x86_64-gcc4.4.7/include/JANA/JEventLoop.h:590
0000019 0x000000000097d943 in DFDCHit_factory::LoadPackageCalibTables (this=0x7f4b6c0f9830, eventLoop=0x7f4b6c0008c0, ccdb_prefix="/FDC/package3") at libraries/FDC/DFDCHit_factory.cc:312
0000020 0x000000000097fca7 in DFDCHit_factory::brun (this=0x7f4b6c0f9830, eventLoop=0x7f4b6c0008c0, runnumber=<value optimized out>) at libraries/FDC/DFDCHit_factory.cc:88
0000021 0x00000000006873c7 in jana::JFactory<DFDCHit>::Get (this=0x7f4b6c0f9830, d=std::vector of length 0, capacity 0) at /home/mstaib/gluex/trunk/jana_0.7.2/Linux_RHEL6-x86_64-gcc4.4.7/include/JANA/JFactory.h:258
0000022 0x0000000000695745 in jana::JEventLoop::GetFromFactory<DFDCHit> (this=0x7f4b6c0008c0, t=std::vector of length 0, capacity 0, tag=0xc4910a "", data_source=
0x7f4b89363290, allow_deftag=<value optimized out>) at /home/mstaib/gluex/trunk/jana_0.7.2/Linux_RHEL6-x86_64-gcc4.4.7/include/JANA/JEventLoop.h:453
0000023 0x0000000000695a25 in jana::JEventLoop::Get<DFDCHit> (this=0x7f4b6c0008c0, t=std::vector of length 0, capacity 0, tag=0xc4910a "", allow_deftag=true) at /home/mstaib/gluex/trunk/jana_0.7.2/Linux_RHEL6-x86_64-gcc4.4.7/include/JANA/JEventLoop.h:304
0000024 0x0000000000989496 in DFDCPseudo_factory::evnt (this=0x7f4b6c0f9b10, eventLoop=0x7f4b6c0008c0, eventNo=<value optimized out>) at libraries/FDC/DFDCPseudo_factory.cc:212
0000025 0x00000000008a30ab in jana::JFactory<DFDCPseudo>::Get (this=0x7f4b6c0f9b10, d=std::vector of length 0, capacity 0) at /home/mstaib/gluex/trunk/jana_0.7.2/Linux_RHEL6-x86_64-gcc4.4.7/include/JANA/JFactory.h:275
0000026 0x00000000008a3c65 in jana::JEventLoop::GetFromFactory<DFDCPseudo> (this=0x7f4b6c0008c0, t=std::vector of length 0, capacity 0, tag=0xc4910a "", data_source=
0x7f4b89363630, allow_deftag=<value optimized out>) at /home/mstaib/gluex/trunk/jana_0.7.2/Linux_RHEL6-x86_64-gcc4.4.7/include/JANA/JEventLoop.h:453
0000027 0x00000000008a3f51 in jana::JEventLoop::Get<DFDCPseudo> (this=0x7f4b6c0008c0, t=std::vector of length 0, capacity 0, tag=0xc4910a "", allow_deftag=true) at /home/mstaib/gluex/trunk/jana_0.7.2/Linux_RHEL6-x86_64-gcc4.4.7/include/JANA/JEventLoop.h:304
0000028 0x00000000009907b9 in DFDCSegment_factory::evnt (this=0x7f4b6c0f9f30, eventLoop=<value optimized out>, eventNo=<value optimized out>) at libraries/FDC/DFDCSegment_factory.cc:91
0000029 0x000000000089d27f in jana::JFactory<DFDCSegment>::Get (this=0x7f4b6c0f9f30, d=std::vector of length 0, capacity 0) at /home/mstaib/gluex/trunk/jana_0.7.2/Linux_RHEL6-x86_64-gcc4.4.7/include/JANA/JFactory.h:275
0000030 0x000000000089db75 in jana::JEventLoop::GetFromFactory<DFDCSegment> (this=0x7f4b6c0008c0, t=std::vector of length 0, capacity 0, tag=<value optimized out>, data_source=
0x7f4b89363940, allow_deftag=<value optimized out>) at /home/mstaib/gluex/trunk/jana_0.7.2/Linux_RHEL6-x86_64-gcc4.4.7/include/JANA/JEventLoop.h:453
0000031 0x000000000089de15 in jana::JEventLoop::Get<DFDCSegment> (this=0x7f4b6c0008c0, t=std::vector of length 0, capacity 0, tag=0xc4910a "", allow_deftag=true) at /home/mstaib/gluex/trunk/jana_0.7.2/Linux_RHEL6-x86_64-gcc4.4.7/include/JANA/JEventLoop.h:304
0000032 0x000000000089aae2 in DTrackCandidate_factory_FDCCathodes::evnt (this=0x7f4b6c108a30, loop=<value optimized out>, eventnumber=<value optimized out>) at libraries/TRACKING/DTrackCandidate_factory_FDCCathodes.cc:152
0000033 0x0000000000689d0b in jana::JFactory<DTrackCandidate>::Get (this=0x7f4b6c108a30, d=std::vector of length 0, capacity 0) at /home/mstaib/gluex/trunk/jana_0.7.2/Linux_RHEL6-x86_64-gcc4.4.7/include/JANA/JFactory.h:275
0000034 0x00000000006940a5 in jana::JEventLoop::GetFromFactory<DTrackCandidate> (this=0x7f4b6c0008c0, t=std::vector of length 0, capacity 0, tag=0xc2d21b "FDCCathodes", data_source=
0x7f4b89363dc0, allow_deftag=<value optimized out>) at /home/mstaib/gluex/trunk/jana_0.7.2/Linux_RHEL6-x86_64-gcc4.4.7/include/JANA/JEventLoop.h:453
0000035 0x0000000000694385 in jana::JEventLoop::Get<DTrackCandidate> (this=0x7f4b6c0008c0, t=std::vector of length 0, capacity 0, tag=0xc2d21b "FDCCathodes", allow_deftag=true) at /home/mstaib/gluex/trunk/jana_0.7.2/Linux_RHEL6-x86_64-gcc4.4.7/include/JANA/JEventLoop.h:304
0000036 0x00000000008426fe in DTrackCandidate_factory::evnt (this=<value optimized out>, loop=0x7f4b6c0008c0, eventnumber=<value optimized out>) at libraries/TRACKING/DTrackCandidate_factory.cc:223
0000037 0x0000000000689d0b in jana::JFactory<DTrackCandidate>::Get (this=0x7f4b6c1061a0, d=std::vector of length 0, capacity 0) at /home/mstaib/gluex/trunk/jana_0.7.2/Linux_RHEL6-x86_64-gcc4.4.7/include/JANA/JFactory.h:275
0000038 0x00000000006940a5 in jana::JEventLoop::GetFromFactory<DTrackCandidate> (this=0x7f4b6c0008c0, t=std::vector of length 0, capacity 0, tag=0xc4910a "", data_source=
0x7f4b89364190, allow_deftag=<value optimized out>) at /home/mstaib/gluex/trunk/jana_0.7.2/Linux_RHEL6-x86_64-gcc4.4.7/include/JANA/JEventLoop.h:453
0000039 0x0000000000694385 in jana::JEventLoop::Get<DTrackCandidate> (this=0x7f4b6c0008c0, t=std::vector of length 0, capacity 0, tag=0xc4910a "", allow_deftag=true) at /home/mstaib/gluex/trunk/jana_0.7.2/Linux_RHEL6-x86_64-gcc4.4.7/include/JANA/JEventLoop.h:304
0000040 0x000000000092b8f5 in DTrackWireBased_factory::evnt (this=0x7f4b6c105e50, loop=0x7f4b6c0008c0, eventnumber=3) at libraries/TRACKING/DTrackWireBased_factory.cc:182
0000041 0x000000000068ad81 in jana::JFactory<DTrackWireBased>::Get (this=0x7f4b6c105e50, d=std::vector of length 0, capacity 0) at /home/mstaib/gluex/trunk/jana_0.7.2/Linux_RHEL6-x86_64-gcc4.4.7/include/JANA/JFactory.h:275
0000042 0x0000000000693555 in jana::JEventLoop::GetFromFactory<DTrackWireBased> (this=0x7f4b6c0008c0, t=std::vector of length 0, capacity 0, tag=0xc4910a "", data_source=
0x7f4b89364480, allow_deftag=<value optimized out>) at /home/mstaib/gluex/trunk/jana_0.7.2/Linux_RHEL6-x86_64-gcc4.4.7/include/JANA/JEventLoop.h:453
0000043 0x0000000000693835 in jana::JEventLoop::Get<DTrackWireBased> (this=0x7f4b6c0008c0, t=std::vector of length 0, capacity 0, tag=0xc4910a "", allow_deftag=true) at /home/mstaib/gluex/trunk/jana_0.7.2/Linux_RHEL6-x86_64-gcc4.4.7/include/JANA/JEventLoop.h:304
0000044 0x0000000000924151 in DTrackTimeBased_factory::evnt (this=0x7f4b6c105fa0, loop=0x7f4b6c0008c0, eventnumber=<value optimized out>) at libraries/TRACKING/DTrackTimeBased_factory.cc:214
0000045 0x00000000005b9c6f in jana::JFactory<DTrackTimeBased>::Get (this=0x7f4b6c105fa0, d=std::vector of length 0, capacity 0) at /home/mstaib/gluex/trunk/jana_0.7.2/Linux_RHEL6-x86_64-gcc4.4.7/include/JANA/JFactory.h:275
0000046 0x00000000005ba3f5 in jana::JEventLoop::GetFromFactory<DTrackTimeBased> (this=0x7f4b6c0008c0, t=std::vector of length 0, capacity 0, tag=<value optimized out>, data_source=
0x7f4b893648b0, allow_deftag=<value optimized out>) at /home/mstaib/gluex/trunk/jana_0.7.2/Linux_RHEL6-x86_64-gcc4.4.7/include/JANA/JEventLoop.h:453
0000047 0x00000000005ba695 in jana::JEventLoop::Get<DTrackTimeBased> (this=0x7f4b6c0008c0, t=std::vector of length 0, capacity 0, tag=0x7f4b9c3c79ff "", allow_deftag=true) at /home/mstaib/gluex/trunk/jana_0.7.2/Linux_RHEL6-x86_64-gcc4.4.7/include/JANA/JEventLoop.h:304
0000048 0x00007f4b9c3507fd in JEventProcessor_commissioning_reactions::evnt (this=0x1e43f30, loop=0x7f4b6c0008c0, eventnumber=3) at JEventProcessor_commissioning_reactions.cc:160
0000049 0x0000000000c0237c in jana::JEventLoop::OneEvent (this=0x7f4b6c0008c0) at src/JANA/JEventLoop.cc:588
0000050 0x0000000000c028f3 in jana::JEventLoop::Loop (this=0x7f4b6c0008c0) at src/JANA/JEventLoop.cc:452
0000051 0x0000000000bcdfcc in LaunchThread (arg=0x7fffb6ccf5f0) at src/JANA/JApplication.cc:1199
#52 0x00000032b6c079d1 in start_thread () from /lib64/libpthread.so.0
#53 0x00000032b68e89dd in clone () from /lib64/libc.so.6
There are no notes attached to this issue.





View Issue Details
439 [Hall D Online] DAQ minor always 2014-12-16 09:51 2014-12-16 09:51
davidl  
 
normal  
new  
open  
none    
none  
   
No
0
0
Makefile dependencies not correct
Sergey mentioned that when one does a "make" in the vmd/src/rol_1 directory that they need to do a "make clean" first because the dependencies are not correct in the file. This should be fixed so that make can work properly.
There are no notes attached to this issue.





View Issue Details
438 [Hall D Online] Databases minor have not tried 2014-12-15 07:23 2014-12-15 07:23
davidl  
 
normal  
new  
open  
none    
none  
   
No
0
0
Bad values in run info DB
The run info DB linked as "Run info vers. 1" on the wiki has numerous entires where either the times are wrong or the first event number os wrong. Look at the individual files for run 2141 for several examples.
There are no notes attached to this issue.





View Issue Details
437 [Hall D Online] Monitoring minor always 2014-12-12 10:52 2014-12-12 10:52
davidl  
 
normal  
new  
open  
none    
none  
   
No
0
0
Monitoring processes fail to start
The monitoring and hdview2 processes are failing to read the correct ET configuration info from COOL.
There are no notes attached to this issue.





View Issue Details
436 [Hall D Online] Computers minor N/A 2014-12-11 20:41 2014-12-11 20:41
davidl  
 
normal  
new  
open  
none    
none  
   
No
0
0
Add disk space alarms to /gluex and gluonfs
Suggestion from Hovanes:

Include reporting of disk space on /gluex and the gluonfs partitions to the raid monitoring system.
There are no notes attached to this issue.





View Issue Details
434 [Hall D Online] Computers tweak N/A 2014-12-10 06:50 2014-12-10 06:50
davidl  
 
low  
new  
open  
none    
none  
   
No
0
0
Send e-mail when RAID disk starts to get filled
We currently have an EPICS alarm go off when the RAID disk gets filled to a certain level. The alarm must then be addressed by someone cleaning out space the disk. It may be useful for an e-mail to be sent at a slightly lower level to some people so that they have a little more time to clear out space before alarms
start going off in the counting house.
There are no notes attached to this issue.





View Issue Details
433 [Hall D Offline] General minor have not tried 2014-12-05 13:45 2014-12-05 13:45
shepherd  
shepherd  
low  
assigned  
open  
none    
none  
   
Problem in dealing with hit sharing in FCAL?
While developing pi0 calibration, Jon Zarling noticed some instances where the sum of the energies of the hits in Cluster does not equal the energy of the cluster. The deviation is minor: < 1 MeV, but I don't believe this should happen. It seems to be a rare occurrence.
There are no notes attached to this issue.





View Issue Details
432 [Hall D Online] General minor have not tried 2014-12-04 13:42 2014-12-04 20:14
sdobbs  
davidl  
normal  
resolved  
fixed  
none    
none  
   
No
0
0
WINWIDTH not being set correctly in Df125Config?
Looking at a recent run which contains Df125Config, the WINWIDTH field is not being set. This field is being used for pedestal subtraction in the DCDCHit_factory class, which is causing the pedestal-subtracted integrals to be incorrect/

An example event is shown in the "additional information field".

Should NSA_NSB be used for pedestal subtraction instead?
From /mss/halld/RunPeriod-2014-10/rawdata/Run001628/hd_rawdata_001628_000.evio
Event 0000003

JANA >>Df125Config:
JANA >> rocid: slot_mask: NSA: NSB: NSA_NSB: NPED: WINWIDTH:
JANA >>---------------------------------------------------------
JANA >> 53 0x07e7f8 65535 65535 43 1 65535
JANA >> 56 0x0fe7f8 65535 65535 43 1 65535
JANA >> 55 0x07e7f8 65535 65535 43 1 65535
JANA >> 62 0x07e7f8 65535 65535 43 1 65535
JANA >> 52 0x07e7f8 65535 65535 43 1 65535
JANA >> 58 0x07e7f8 65535 65535 43 1 65535
JANA >> 57 0x0fe7f8 65535 65535 43 1 65535
JANA >> 61 0x07e7f8 65535 65535 43 1 65535
JANA >> 60 0x0fe7f8 65535 65535 43 1 65535
JANA >> 59 0x0fe7f8 65535 65535 43 1 65535
JANA >> 28 0x03e7f8 65535 65535 85 1 65535
JANA >> 25 0x03e7f8 65535 65535 85 1 65535
JANA >> 26 0x01e7f8 65535 65535 85 1 65535
JANA >> 27 0x01e7f8 65535 65535 85 1 65535
JANA >>
Notes
(0000623)
davidl   
2014-12-04 20:14   
This was due to an unannounced change that switched from using WINWIDTH for the f125 to NSA_NSB in order to bring it into alignment with how the f250 is being handled.

Incidentally, the handling of this config parameter is now done automatically in the DAQ plugin where the value is copied into nsamples_integral member of the PulseIntegral and DigiHit objects.





View Issue Details
430 [Hall D Offline] General minor always 2014-12-03 16:14 2014-12-04 20:08
davidl  
davidl  
normal  
resolved  
fixed  
none    
none  
   
Number of pedestal samples reported by emulation is 4 when it should be 1
It appears that the number of samples being reported in the Digi Hit objects when generated from emulated data is 4 when it should be 1. Simon pointed this out when looking at ST data.
Notes
(0000622)
davidl   
2014-12-04 20:08   
This was fixed in svn revision 16752. Here is the e-mail announcing it:

Hi All,

  I have just checked in some changes to the commissioning branch that will hopefully help
improve the situation regarding number of samples used to calculate pedestals to be subtracted
from integrals. These include:

1.) Adding nsamples_integral and nsamples_pedestal fields to the Df125PulseIntegral class as
   well as the DCDCDigiHit and DFDCCathodeDigiHit classes.

2.) Fixed the problem where the value of samples_pedestal was always set to 4 for the Df250PulseIntegral
   objects. Now it is set to 1.

3.) The value of nsamples_integral in both the Df1250PulseIntegral and Df250PulseIntegral objects
   are now set automatically to NSA_NSB iff the values came from mode7 data and a config. object
   was found in the data stream. Otherwise, it is set to what was used in emulation so in all cases, it should
   be valid.


So, for people implementing code that performs pedestal subtraction, you should do something like
the following:

double integral = (double)pulse_integral->integral;
double single_sample_ped = (double)pulse_integral->pedestal;
double nsamples_integral = (double)pulse_integral->nsamples_integral;
double nsamples_pedestal = (double)pulse_integral->nsamples_pedestal;
double pedestal = single_sample_ped * nsamples_integral/nsamples_pedestal;
double dE = gain * (integral - pedestal);


i.e. you should no longer reference the NSA_NSB from the config objects directly. This mechanism should
work for both emulated and non-emulated pulse integral objects. Let me know if anyone sees any issues.

Regards,
-David





View Issue Details
429 [Hall D Offline] General minor always 2014-12-03 16:10 2014-12-04 08:01
shepherd  
davidl  
low  
resolved  
fixed  
none    
none  
   
type mismatch
I notice there is some type mismatch between the Df250PulseIntegral class and, say, the DFCALDigiHit class. (It probably exists in other DigiHit classes also.)

In DTranslationTable.h class there is this template:

template<class T>
void DTranslationTable::CopyDf250Info(T *h, const Df250PulseIntegral *pi, const Df250PulseTime *pt, const Df250PulsePedestal *pp) const
{
    /// Copy info from the fADC250 into a hit object.
    h->pulse_integral = pi->integral;

For the last assignment the RHS is of type int32_t, while the LHS is of type uint32_t. It turns out the RHS is initialized from data of type uint32_t in the constructor of the Df250PulseIntegral object. Therefore, I *think* this is harmless, but it would be good to correct it for consistency.
Notes
(0000621)
davidl   
2014-12-04 08:01   
Types of both integral and pedestal members of Df250PulseIntegral and Df125PulseIntegral to be uint32_t in revision 16742.





View Issue Details
428 [Hall D Offline] General major always 2014-12-03 16:00 2014-12-04 07:21
shepherd  
davidl  
urgent  
resolved  
fixed  
none    
none  
   
Random variation of integral values in FCALDigiHits
For this file:

/cache/mss/halld/RunPeriod-2014-10/rawdata/Run001518/hd_rawdata_001518_000.evio

Running hd_dump produces this output for FCALDigiHit:

================================================================
Event: 5

JANA >>DFCALDigiHit:
JANA >> row: column: pulse_integral: pulse_time: pedestal: QF:
JANA >>--------------------------------------------------------
JANA >> 27 34 5066 733 101 1
JANA >> 29 45 6346 933 101 1
JANA >> 40 55 6386 785 101 1
JANA >> 29 55 5105 6144 102 1
JANA >> 13 41 5038 2592 100 1
JANA >> 13 44 5526 608 100 1
JANA >> 47 36 6393 814 100 1
JANA >> 52 36 5016 2304 100 1
JANA >> 30 13 4982 3557 100 1
JANA >> 14 20 5147 2402 100 1
JANA >> 27 2 4376 1005 101 1
JANA >> 27 3 5624 1385 102 1
JANA >> 0 23 5472 1352 99 1
JANA >> 0 22 4329 1326 101 1
JANA >> 39 2 5007 4068 99 1
JANA >> 57 20 5041 1962 101 1
JANA >> 57 25 5287 1128 100 1
JANA >> 21 52 5560 3934 100 1 <<--- compare this one

A stand-alone plugin run with hd_root produces this with columns: row column integral time samples_integral.

Event: 5
27 34 5066 733 50
29 45 6346 933 55
40 55 6386 785 55
29 55 5105 6144 50
13 41 5038 2592 50
13 44 5526 608 55
47 36 6393 814 55
52 36 5016 2304 50
30 13 4982 3557 50
14 20 5147 2402 51
27 2 4376 1005 50
27 3 5624 1385 55
0 23 5472 1352 55
0 22 4329 1326 50
39 2 5007 4068 50
57 20 5041 1962 50
57 25 5287 1128 53
21 52 61896 3934 45 <<-- to this one

Note the last hit has a different integral. This is seen in multiple events:

================================================================
Event: 6
JANA >>DFCALDigiHit:
JANA >> row: column: pulse_integral: pulse_time: pedestal: QF:
JANA >>--------------------------------------------------------
JANA >> 27 40 6421 935 100 0
JANA >> 29 43 5024 1211 101 1
JANA >> 43 51 5046 5158 100 1
JANA >> 28 51 5024 5621 100 1
JANA >> 29 52 5015 3640 100 1
JANA >> 13 44 5221 473 101 1
JANA >> 34 36 5513 6037 99 1
JANA >> 41 40 4581 0 0 1
JANA >> 37 43 7426 4515 99 1
JANA >> 46 35 5596 2916 101 1
JANA >> 41 44 5004 494 103 1
JANA >> 37 46 4982 546 100 1
JANA >> 33 27 5077 5300 101 1
JANA >> 37 19 5007 5794 100 1
JANA >> 14 18 6386 810 100 1
JANA >> 22 12 28485 5166 100 1 <<--- compare this one
JANA >> 14 11 5083 373 102 1
JANA >> 24 22 5077 1111 101 1
JANA >> 21 21 6399 682 101 1
JANA >> 28 12 5092 445 100 1
JANA >> 30 58 5025 5742 100 1

Event: 6
27 40 6421 935 55
29 43 5024 1211 50
43 51 5046 5158 50
28 51 5024 5621 50
29 52 5015 3640 50
13 44 5221 473 52
34 36 5513 6037 55
41 40 4581 0 53
37 43 3514 4515 36
46 35 5596 2916 55
41 44 5004 494 50
37 46 4982 546 50
33 27 5077 5300 50
37 19 5007 5794 50
14 18 6386 810 55
22 12 2305 5166 26 <<-- to this one
14 11 5083 373 50
24 22 5077 1111 50
21 21 6399 682 55
28 12 5092 445 51
30 58 5025 5742 50

If this is a low level bug, it could have widespread effects. There is some
chance the standalone code has a bug that is corrupting memory, but
it is difficult to see how this is possible as it is just a loop over DFCALDigiHit
objects.

Notes
(0000620)
davidl   
2014-12-04 07:21   
Matt Shepherd identified the problem as adding in one sample past the end of the array for late hits when emulating the Df250PulseIntegral objects. A fix has been checked in to the commissioning branch in revision 16738 on 12/3/2014





View Issue Details
431 [Hall D Offline] General minor N/A 2014-12-03 16:19 2014-12-03 16:19
davidl  
 
normal  
new  
open  
none    
none  
   
fADC emulation needs to be tailored by detector.
The fADC emulation needs to be tailored by detector to match the appropriate NSA, NSB setting for that detector. This should be done using the associated Df250Config object since we would otherwise need to access the translation table inside the DAQ plugin to figure out which detector a hit belonged to.

The main issue will be how to determine NSA and NSB individually from the single number NSA_NSB reported in the config object. We may just need to use a map that converts from specific NSA_NSB values to known NSA, NSB values.
There are no notes attached to this issue.





View Issue Details
355 [Hall D Online] Databases feature N/A 2013-06-14 10:15 2014-12-03 06:04
wolin  
sergpozd  
normal  
assigned  
open  
none    
none  
   
No
3
0
0
Improve history mechanism in JInventory
Currently history is just a concatenation of changes in a text field. Better would be a separate history table indexed by item, date and perhaps other categories.
Separate table and database/php code is simple, but good design needed for gui.
There are no notes attached to this issue.





View Issue Details
422 [Hall D Online] Monitoring major always 2014-12-02 15:42 2014-12-03 06:03
davidl  
davidl  
high  
assigned  
open  
none    
none  
   
No
0
0
RootSpy not working properly in online mode
The RootSpy GUI has not been working correctly for the run up to now. It does not appear to be properly remembering histogram selections. It also seems to be unable to get histograms during normal data
running. Sometimes behaving as though the processing nodes are not sending their histograms in response to requests.

The program needs to be re-tested and debugged in the counting house environment to make it useful for shift takers.
There are no notes attached to this issue.





View Issue Details
425 [Hall D Offline] General minor have not tried 2014-12-02 17:01 2014-12-03 06:03
shepherd  
shepherd  
normal  
assigned  
open  
none    
none  
   
Isolated crash in DFCALCluster_factory
Reported by Paul:
We saw at least one job crash while running the monitoring plugins, and it occurred in DFCALCluster::evnt():

/volatile/halld/RunPeriod-2014-10/offline_monitoring/001156/2014-11-22/log/stderr_001156_000.err

Could you (or someone in the FCAL group) take a look at this when you get the chance? This was while running over file 0 of run 1156.
The stack trace references libraries/FCAL/DFCALCluster_factory.cc:189

In this loop there are some pointer dereferences, but it is not clear how any of them could be bad pointers.

This needs verification by rerunning the job.

There are no notes attached to this issue.





View Issue Details
339 [Hall D Offline] General trivial always 2013-05-03 20:56 2014-12-02 16:49
pmatt  
davidl  
low  
resolved  
fixed  
none    
none  
   
Minor EVENTS_TO_SKIP bug
In the 0.6.6p1 release of JANA, on line 559 of JApplication.cc (in the EventBufferThread function), the EVENTS_TO_SKIP condition should be <= instead of <. You can see this by setting EVENTS_TO_SKIP = 1 and it won't skip anything. It's because NEvents_Read is incremented to 1 in ReadEvent, prior to this check.
This is ridiculously minor, and I would check it in myself if I knew how.
Notes
(0000618)
davidl   
2014-12-02 16:49   
I believe this was fixed quite some time ago, but never marked as resolved.





View Issue Details
332 [Hall D Online] Networking feature N/A 2013-03-27 11:56 2014-12-01 11:00
wolin  
davidl  
normal  
resolved  
fixed  
none    
none  
   
Yes
4
0
0
Specify, purchase and install InfiniBand hardware and software
Purchase InfiniBand hardware as part of CC large purchase May-2013, install drivers, configure IB network
Dave Abbott from the DAQ group will do this
Notes
(0000617)
davidl   
2014-12-01 11:00   
Done.





View Issue Details
258 [Hall D Online] Networking feature N/A 2012-04-24 17:29 2014-12-01 10:58
wolin  
davidl  
normal  
resolved  
fixed  
none    
none  
   
No
3
0
0
Develop scheme to remotely monitor and reset network switches
In order to avoid shutting down the beam unnecessarily we need a remote reset facility for network switches (and other components). The mechanism should not rely on the network being operational. Bryan Hess says that cycling the power is the best way to reset a network switch.
Bryan has an extensive network monitoring and alarm system, we probably should use it.
Notes
(0000614)
davidl   
2014-12-01 10:58   
IT division has provided good support for this. It is no longer clear that we need a separate facility for us to manage the switches.





View Issue Details
155 [Hall D Online] EPICS feature N/A 2011-10-24 11:41 2014-12-01 10:52
wolin  
davidl  
normal  
resolved  
fixed  
none    
none  
   
No
4
0
0
Survey EPICS backup/restore systems and choose one
Survey EPICS backup and restore facilities, test and make recommendation.
Notes
(0000500)
wolin   
2013-03-21 14:00   
Not sure who will do this.
(0000610)
davidl   
2014-12-01 10:52   
Hovanes has taken care of this.





View Issue Details
239 [Hall D Online] Monitoring feature N/A 2012-04-10 11:50 2014-12-01 10:51
wolin  
davidl  
normal  
resolved  
fixed  
none    
none  
   
No
6
0
0
Incorporate CLAS mini-Hbook into RootSpy
Upgrade RootSpy to receive and process histograms in mini-Hbook format, the lightweight format used by CLAS to collect histogram info in the front-end ROC's. Design mechanism to transport histograms from ROC's to RootSpy.
cMsg is the obvious choice for transport.
Notes
(0000375)
wolin   
2012-05-30 17:11   
CNU summer student Rob Ryan will be looking into this.
(0000501)
wolin   
2013-03-21 14:01   
Not sure who will do this.
(0000609)
davidl   
2014-12-01 10:51   
This was done by CNU student (Rob) some time ago.





View Issue Details
317 [Hall D Online] DAQ feature N/A 2013-03-21 17:26 2014-12-01 10:49
wolin  
davidl  
normal  
resolved  
fixed  
none    
none  
   
No
6
0
0
Front end readout lists
Develop readout list system for all front-end crates.
Notes
(0000607)
davidl   
2014-12-01 10:49   
Initial working versions were developed for the Fall 2014 engineering run.





View Issue Details
333 [Hall D Online] DAQ feature N/A 2013-03-27 11:59 2014-11-30 08:52
wolin  
davidl  
normal  
resolved  
fixed  
none    
none  
   
Yes
10
0
0
Configure, test and install single board computers
Setup and test single board computers, install in crates, develop boot system on HDCC, install DHCP system if needed.
Bryan Moffit will do this.
Notes
(0000605)
davidl   
2014-11-30 08:52   
Done (long ago)





View Issue Details
154 [Hall D Online] Controls feature N/A 2011-10-24 11:40 2014-11-30 08:52
wolin  
davidl  
normal  
resolved  
fixed  
none    
none  
   
No
8
0
0
Survey EPICS alarm systems, choose one and implement
Survey and test EPICS alarm systems, choose and implement. This will be the central alarm system for Hall D.
Important criteria involve ability of the system to deal with alarms generated from other than EPICS, i.e. we would like this to be the main alarm system for the entire experiment, accepting alarms from all online systems, whether EPICS or not.

Currently BEAST from SNS is the leading contender
Notes
(0000209)
hovanes   
2011-11-16 09:06   
Start with evaluating BEAST
(0000499)
wolin   
2013-03-21 14:00   
Not sure who will do this.
(0000526)
wolin   
2013-06-13 14:24   
Hovanes is testing BEAST, part of the CSS suite.
(0000604)
davidl   
2014-11-30 08:52   
It looks like an alarm system has been chosen and implemented.





View Issue Details
92 [Hall D Online] DAQ feature N/A 2011-06-17 15:57 2014-11-30 08:40
wolin  
davidl  
normal  
resolved  
fixed  
none    
none  
   
Yes
8
0
0
Design disentangling architecture and develop prototype event disentangler
Design architecture appropriate for event disentangling in the Hall D DAQ system. Design a CODA event building component that disentangles events prior to the farm system and event recorder such that event data is contiguous in memory. Software should include working example that disentangles data from F1TDC, fADC125 and fADC250 modules that will be used in Hall-D. Develop and test prototype in realistic environment. Deliver report that summarizes results, recommends a production architecture and includes an outline of steps needed to create and commission production disentangler with estimates of required time and manpower.
Consult as required with the Hall D Online group concerning the specific requirements and constraints of the Hall D computing system. Major architectural decisions must be reviewed by the Hall D Online group to ensure compatibility with the experiment online environment.
Notes
(0000120)
wolin   
2011-06-20 10:16   
Graham...kindly review the task description and modify as required, especially the estimated man-weeks of effort needed. We'll iterate as needed.
(0000121)
wolin   
2011-06-20 10:17   
Please update task description as needed
(0000148)
heyes   
2011-08-31 07:28   
"Software should include working example that disentangles data from F1TDC, fADC125 and fADC250 modules that will be used in Hall-D".

It isn't clear that this is possible at this time since the data format from the various modules isn't stable yet. A prototype "proof of principle" is possible.
(0000251)
wolin   
2012-01-23 10:04   
Probably can disentangle in quad-core or better ROC or in EMU stage. DAQ group will determine how much CPU is available in ROC at high-rates.
(0000478)
wolin   
2013-03-19 11:56   
I believe Carl has made progress on this. Different strategies for where to run the disentangler (ROC, EB, both?) will be investigated in the Hall D DAQ test setup by the DAQ group.
(0000503)
wolin   
2013-03-21 14:04   
Not sure who will do this.
(0000556)
wolin   
2013-07-18 10:12   
Dave and I now realize the two best places to do disentangling are on the ROC and in the L3 farm. I.e. a separate disentangling EB stage may not be the best choice. We need to further investigate these two strategies.
(0000600)
davidl   
2014-11-30 08:40   
Event disentangling is currently being handled in the DAQ plugin in the Hall-D software. It will likely stay there since it allows the disentangling job to be distributed to multiple computers rather than be done on a single computer if it were handled in the event builder.





View Issue Details
70 [Hall D Online] Computers feature N/A 2011-04-12 17:34 2014-11-30 08:38
wolin  
davidl  
normal  
resolved  
fixed  
none    
none  
   
No
2
0
0
Evaluate computer and process monitoring systems
We need to monitor the health of computers and processes/services on those computers, and need a facility for starting processes when they fail. Nagios and Ganglia are leading candidates. The CC uses Ganglia+Mon, CLAS uses Nagios.
Notes
(0000084)
wolin   
2011-04-15 15:23   
(edited on: 2011-04-15 15:24)
Current leading candidates: Ganglia, Nagios, Mon, Zabbix and OpenNMS.

(0000567)
wolin   
2013-07-26 16:31   
CC is familiar with Ganglia, considering moving to something else. Might not know for a while. Paul will install Ganglia to get us going.
(0000599)
davidl   
2014-11-30 08:38   
Paul Letta installed Ganglia which is available from the internal network on gluonweb.jlab.org





View Issue Details
166 [Hall D Online] Databases feature N/A 2011-11-15 11:30 2014-11-30 08:37
wolin  
davidl  
normal  
resolved  
fixed  
none    
none  
   
No
6
0
0
Develop unified front-end board configuration strategy and database
All front-end boards need to be configured, both for trigger and readout purposes. A unified system is desirable. Interoperation with the conditions database needs to be considered as well.
It is not clear if a single system could be used for all configurations. The EPICS backup/restore facility naturally should handle EPICS configurations. It is further not clear how to configure our many HV and LV channels. Again, a unified system is desirable, if possible.
Notes
(0000598)
davidl   
2014-11-30 08:37   
Alex S. has implemented a system using text files to configure all front end boards.





View Issue Details
312 [Hall D Online] Monitoring feature N/A 2013-03-21 14:07 2014-11-30 08:35
wolin  
davidl  
normal  
resolved  
fixed  
none    
none  
   
No
8
0
0
Develop JANA-based farm process
JANA-based processes will run on the monitoring/L3 farm. A JANA variant suitable for online use must be designed and written.
E.g. it must incorporate cMsg and ET.
Notes
(0000529)
wolin   
2013-06-13 14:27   
Vardan and I agreed to start by using the CodaObject framework for command/control. This will need to be incorporated into hd_online, along with cMsg, evio, et, etc.
(0000597)
davidl   
2014-11-30 08:35   
This was done some time ago (hdmon program).





View Issue Details
325 [Hall D Online] Monitoring feature N/A 2013-03-22 10:15 2014-11-30 08:35
wolin  
davidl  
normal  
resolved  
fixed  
none    
none  
   
No
3
0
0
Implement translation table in JANA
Complete design of translation table, implement in JANA.
Notes
(0000596)
davidl   
2014-11-30 08:35   
This was done quite some time ago but never marked as resolved.





View Issue Details
419 [JANA] Feature Request minor N/A 2014-09-29 07:39 2014-11-10 10:17
davidl  
davidl  
normal  
resolved  
fixed  
none    
none  
   
create jview plugin
A nice tool might be to have a GUI program that allows one to interactively explore an event's objects. It would allow one to click on an object to activate that factory and to automatically display the call graph. Inspection of values and associated objects could be done as well. Clicking an associated object would bring it up along with its call graph, list of it's associated objects as well as a list of objects for which it is
itself and associate object.

It may be that this could actually be done as a plugin so that it could be written in generic JANA, but used by hd_ana.
Notes
(0000595)
davidl   
2014-11-10 10:17   
This has been implemented as the janaview plugin in JANA. It uses ROOT for the GUI seems to be working OK. It will appear in JANA 0.7.3





View Issue Details
414 [JANA] Bug minor always 2014-08-14 22:10 2014-08-14 22:13
davidl  
davidl  
normal  
resolved  
fixed  
none    
none  
   
Program exists if hi-res timer runs out
Aristeidis reported this:

For some especially long jobs (ones running for more than 1 day) the program would suddenly die with a message about a timer running out. This turns out to be due signals being sent when any of the 3 hi-res timers we use runs out. The default behavior for uncaught signals of these types is to terminate the program. The timers are already set for 1M seconds, but apparently, with multiple threads, this counts quicker than the wall clock, (Otherwise, they would run for over 270 hours before timing out.)

To fix this, alarm handlers need to be installed for SIGALRM, SIGPROF, and SIGVTARLM signals.
Notes
(0000594)
davidl   
2014-08-14 22:13   
This fix was committed in revision 2075 and will appear in JANA 0.7.2





View Issue Details
413 [JANA] Feature Request minor always 2014-08-14 21:34 2014-08-14 21:35
davidl  
davidl  
normal  
resolved  
fixed  
none    
none  
   
Automatically compare md5 checksum for all resources before using
This was from Matt Shepherd.

If a process was interrupted while downloading a resource, the file can be incomplete. Subsequent invocations of the JANA program will see the partial file and assume it already exists and is complete. This causes confusing behavior as the program can set. fault but not really point to the actual reason. To avoid this, calculated the md5sum for all resources as they are requested and compare them against what is stored in the calib DB. If they don't match, print an informative error message and exit. Give the user an option to skip the checksum if they want.
Notes
(0000593)
davidl   
2014-08-14 21:35   
This was submitted in revision 2074 and will appear in JANA 0.7.2





View Issue Details
412 [JANA] Feature Request minor have not tried 2014-07-18 08:19 2014-07-18 08:30
davidl  
davidl  
normal  
resolved  
fixed  
none    
none  
   
Set pointer to NULL in JEventLoop::GetSingle if 0 values.

From an e-mail from Paul Mattione on June 7, 2014

===========================================
In JEventLoop::GetSingle(), could you set the pointer to NULL if the vector is empty, and only throw an exception if the size is >= 2? (Current is to throw if size != 1). I've been using GetSingle() for DMCThrownMatching, and checking for NULL after the function call, but it throws on (CLAS) experimental data.

Thanks.

- Paul
===========================================
Notes
(0000592)
davidl   
2014-07-18 08:30   
This was fixed in svn revision 2046 and will appear in JANA 0.7.2

The code now sets the value of "t" to NULL in all cases where there is not exactly one object of the requested type. It also has a 3rd parameter that can be used to disable throwing an exception if N!=1.





View Issue Details
398 [JANA] Feature Request minor always 2014-02-18 07:45 2014-07-18 07:50
davidl  
davidl  
normal  
resolved  
fixed  
none    
none  
   
THREAD_TIMEOUT should supersede THREAD_TIMEOUT_FIRST_EVENT if larger
A few people have been bitten by trying to set THREAD_TIMEOUT to a large value only to have it timeout on the first event because THREAD_TIMEOUT_FIRST_EVENT is set to something smaller. If it is smaller, then THREAD_TIMEOUT should be used.
Notes
(0000590)
davidl   
2014-07-18 07:50   
This was fixed in svn revision 2045 and will appear in JANA 0.7.2





View Issue Details
410 [JANA] Feature Request minor N/A 2014-06-04 08:57 2014-07-18 07:42
davidl  
davidl  
normal  
resolved  
fixed  
none    
none  
   
Default directory for JANA plugins + error message
From an e-mail sent by Mauri 4/14/2014:

-------------
Thanks, that works. May I suggest to add "make sure the JANA_PLUGIN_PATH is set correctly" to your error message?

Also, I would suggest that $JANA/plugins is checked always, since those plugins come with the JANA system. JANA_PLUGIN_PATH should be set for user-defined plugins in my opinion.
In other words, we could limit the default number of env variables to one ;-)
-------------

The "error message" he refers to is the following:

JANA ERROR>>
JANA ERROR>>***ERROR : Couldn't find plugin "janadot"!***
JANA ERROR>>*** To see paths checked, set PRINT_PLUGIN_PATHS config. parameter

 
Notes
(0000589)
davidl   
2014-07-18 07:42   
This was fixed in revision 2044 and will appear in JANA 0.7.2

The error message was added as requested.

The $JANA_HOME/plugins directory was added as a default search directory. This was actually already in, but it was still using $JANA_HOME/lib. The plugins directory was added in 0.7.0 when the scons build system was introduced, but the default search directory was not updated.





View Issue Details
402 [Hall D Online] Monitoring minor have not tried 2014-04-08 11:10 2014-07-16 14:05
davidl  
sdobbs  
normal  
assigned  
open  
none    
none  
   
No
0
0
ROOTSpy: Add Macros for complex displays
For the CDC, a convenient way of holding occupancy information is using a single histogram for each ring. (28 histograms total). These need to be plotted all on the same axes and using polar coordinates (the "pol" option.) Complex rendering like this may be needed in other places as well and is probably best handled using a macro supplied by the author.

Two options should be implemented:

1.) The server sends the macro. This allows, in principle, the author of the histogram to send it from a string embedded inside plugin code. It would guarantee the macro was available whenever and wherever the histogram is made.

2.) The client uses a macro or set of macros at a location they specify. This could be useful for setting up shift-taker situations where they are pulling histograms from lots of places and wish to display a subset, possibly with a divided canvas for efficiency.

One possibility is to allow the server to package a tree or set of histograms in a TMemFile object (as is currently done for TTrees). This could allow including macros as well in the form of TObjString objects. The TFileMerger class can be used to combine multiple TMemFiles similar to how histograms are added.
Notes
(0000586)
sdobbs   
2014-04-09 12:55   
I've been thinking that to handle the more complicated displays, it would be helpful to spin off the code that handles drawing the histograms/whatever from rs_mainframe into its own class.

Also, for creating displays for shift-taking, one potentially straightforward way would be to allow the creation of a new window with multiple TPads (one per histogram), as opposed to splitting the TPad. There does exist the ability to link a client side script to be run every time that a histogram is displayed, which allows for limited client-side customizability.

There is also the ability to define sets of histograms that can be activated all at once. It could be possible to piggyback the creation of views for shift takers on top of this.





View Issue Details
399 [Hall D Online] Monitoring minor N/A 2014-04-08 10:30 2014-07-16 12:44
davidl  
sdobbs  
normal  
assigned  
open  
none    
none  
   
No
0
0
ROOTSpy: Allow user to modify window size
The canvas does not currently resize itself when the window size/shape are changed.
Notes
(0000588)
sdobbs   
2014-07-16 12:44   
The canvas and lower UI elements currently resize properly - am working on the upper UI elements.





View Issue Details
411 [Hall D Offline] General feature N/A 2014-06-20 08:15 2014-06-20 08:15
davidl  
 
normal  
new  
open  
none    
none  
   
GetCDCWires and GetFDCWires inefficient
The GetCDCWires and GetFDCWires methods of DGeometry will allocate new wire structures every time they are called. This leads to duplication for each thread and possible memory leaks, For example, the wires are needed in the DCDCHit_factory class, but only to get the number of straws for each ring. The wires then have to be deleted one at a time.

The wires should be cached as something like vector<vector<const DCDCWire*> > and the signatures of GetCDCWires changed to use that as well. Similarly for FDC.
There are no notes attached to this issue.





View Issue Details
397 [Hall D Offline] General minor always 2014-02-05 11:05 2014-06-20 08:09
romanov  
davidl  
normal  
resolved  
fixed  
none    
none  
   
Sim-recon SBMS doesn't generate setup.sh for Bash shell
After calling
scons install

setup.sh for Bash shell is not created in sim-recon/$BMS_OSNAME directory.
Notes
(0000587)
davidl   
2014-06-20 08:09   
Committed in svn revision 11822. (Not thoroughly tested, but there.)





View Issue Details
401 [Hall D Online] Monitoring minor have not tried 2014-04-08 10:50 2014-05-27 18:11
davidl  
sdobbs  
normal  
resolved  
open  
none    
none  
   
No
0
0
ROOTSpy: Reset Button
It would be useful to have a button that can make the RootSpy GUI forget any servers or histograms it currently has record of. This would be most useful when trying to develop software that produces the histograms and one is continually restarting the server. The old histograms in memory will just interfere.

Because this could, in a production environment, lead to mistakes that lose some information, a confirmation dialog might be in order.
This has been implemented as an option in the "Tools" menu, with a confirmation dialog.
There are no notes attached to this issue.





View Issue Details
313 [Hall D Online] Monitoring feature N/A 2013-03-21 14:09 2014-05-27 17:48
wolin  
sdobbs  
normal  
resolved  
open  
none    
none  
   
No
5
0
0
Develop RootSpy archive/retrieval system
Farm processes will generate ROOT historgram, and RootSpy will be used to access them. Develop a RootSpy program that archives and displays histograms.
The RootSpy archiver has been implemented and initial testing completed during the 2013 Online Data Challenge.
There are no notes attached to this issue.





View Issue Details
409 [Hall D Offline] General major have not tried 2014-04-18 10:44 2014-04-18 10:44
marki  
sdobbs  
normal  
assigned  
open  
none    
none  
   
look further into EventStore
from 4/14/16 meeting
There are no notes attached to this issue.





View Issue Details
407 [Hall D Offline] General major have not tried 2014-04-18 10:41 2014-04-18 10:42
marki  
jonesrt  
normal  
assigned  
open  
none    
none  
   
continue running on the Grid: Richard
from 4/16/14 meeting
There are no notes attached to this issue.





View Issue Details
405 [Hall D Offline] General major have not tried 2014-04-18 10:39 2014-04-18 10:40
marki  
marki  
normal  
assigned  
open  
none    
none  
   
copy dc2 events to JLab
from the various sites, put on tape
There are no notes attached to this issue.





View Issue Details
404 [Hall D Online] Monitoring minor have not tried 2014-04-15 19:18 2014-04-15 19:18
sdobbs  
 
normal  
new  
open  
none    
none  
   
No
0
0
RootSpy: Optimize GUI during auto-update
I've noticed that sometimes the GUI can be less responsive when auto-update is turned on. For example, usually button presses or check box selections won't show up until the next update happens. This can frustrating when trying to test settings or flip to other histograms before the next update.

The GUI should either be made more asynchronous, or some of the processing should be done in extra threads.
There are no notes attached to this issue.





View Issue Details
400 [Hall D Online] Monitoring minor N/A 2014-04-08 10:44 2014-04-09 09:44
davidl  
davidl  
normal  
resolved  
fixed  
none    
none  
   
No
0
0
ROOTSpy: Add mutex lock mechanism for coordinating with user
Synchronization issues can (and sometimes do) occur when the RootSpy thread accesses a histogram or tree at the same time as the main program. We need to provide a mechanism by which they can coordinate. This should be a pthread_mutex_t so that it can be coupled easily with the existing mutex used in GlueX sim-recon code.

The DRootSpy constructor should take an optional pthread_mutex_t* that is initialized to NULL if not given. This will allow any existing code to still compile. A global variable:

   pthread_mutex_t *ROOTSPY_MUTEX = NULL;

should be created and set to the value passed to the DRootSpy constructor if present. If not, then a mutex should be allocated and the pointer set to that. In either case the RootSpy thread should then use the pointer to lock and unlock the mutex for ROOT operations.

 Users that don't provide the mutex pointer can automatically make use of the global ROOTSPY_MUTEX if they wish to synchronize. Otherwise, they are taking their own chances, which is what the situation is now!


Notes
(0000585)
davidl   
2014-04-09 09:44   
This was checked in on 4/9/2014 as svn revision 13200.





View Issue Details
379 [Hall D Online] Monitoring feature N/A 2013-09-09 17:18 2014-04-08 11:13
wolin  
wolin  
normal  
resolved  
fixed  
none    
none  
   
No
1
0
0
Create codaObject JANA plugin
Create plugin to implement codaObject functionality in JANA. This will allow JANA processes to be controlled by run control or other similar systems.
Notes
(0000584)
davidl   
2014-04-08 11:13   
Elliott did do this at some level and the code is in the repository. I believe limited testing was done and none by anyone other than Elliott. I'm closing this for now to try and clear out the Mantis CUE.





View Issue Details
396 [Hall D Offline] General feature N/A 2014-02-02 22:51 2014-02-02 23:45
romanov  
 
low  
new  
open  
none    
none  
   
CCDB. Warn user to add data for 0-inf run range
Suppose, a new table is created. So there is no data yet. User adds some data for specific run range (run 0-1000 as an example). CCDB CLI should warn the user that there must be at least one data set that covers all runs (0-inf)

Unit tests of what happens if there is no data in table for current run is a good idea too.
There are no notes attached to this issue.





View Issue Details
136 [Hall D Offline] General minor have not tried 2011-09-21 13:02 2014-02-02 22:45
marki  
romanov  
normal  
resolved  
fixed  
none    
none  
   
Decide how to handle large calibration sets
problem with default MySQL configuration
Notes
(0000583)
romanov   
2014-02-02 22:43   
The issue is closed by using resources and resource manager in sim-recon





View Issue Details
362 [JANA] Bug minor always 2013-06-28 08:16 2014-01-23 09:40
davidl  
davidl  
normal  
resolved  
fixed  
none    
none  
   
null_data_source not defined
This is actually a couple of issues. These are related and came up while implementing the translation table in the DAQ and TTab plugins. The desired behavior was to have the DAQ plugin create low level objects and the TTab plugin take those and generate detector hit objects. If only the DAQ plugin is present, no translation is done. If the TTab plugin is present, then translation is done automatically.

This was to be achieved by simply calling JEventLoop::GetSingle from GetObjects to get a DTranslationTable object that can then be used to generate the hit objects by applying the table (via the ApplyTranslationTable method). The issues that came up were:

1.) When calling JEventLoop::Get from a JEventSource::GetObjects method, an infinite recursion is entered and doesn't stop until an "illegal instruction" is encountered. It would be good to have some way of detecting this for event sources as is done for factories.

2.) The temporary solution was to use GetFromFactory instead of GetSingle. This, however, gave a run time link error saying "null_data_source" was not defined. Giving an explicit argument of JEventLoop::DATA_FROM_FACTORY also doesn't work since the argument requires a reference. A temporary variable must then be defined so a reference to it can be passed. In the end the code looked like this:

JEventLoop::data_source_t type = JEventLoop::DATA_FROM_FACTORY; // (this needs to get fixed in JANA)
loop->GetFromFactory(translationTables, "", type);

when it could have looked like this:
oop->GetFromFactory(translationTables);


Notes
(0000579)
davidl   
2013-12-13 16:23   
Item 1.) here is not easy to implement in a way that won't risk significant performance degradation. For normal factories, the Get method is implemented in the JFactory base class which allows us to use a "busy" flag to check quickly and easily for recursion. For event sources, the first point where we have access by the framework is in the JEventLoop::Get() method. Here, we would have to turn on call stack tracing and check it each time to ensure there is no duplicate entry. We could also put some arbitrary limit on the depth of the call stack without actually recording it, but that would have to be high enough to never be a problem for non-infinite-recursion cases, but be small enough to catch the infinite recursion cases before they caused memory errors.
(0000580)
davidl   
2013-12-13 16:30   
Made data_source by passed by value rather than by reference. Since it is an enum, it really saved almost nothing doing it by reference in the first place.





View Issue Details
389 [Hall D Offline] General minor have not tried 2013-12-17 14:12 2014-01-06 09:46
marki  
staylor  
normal  
resolved  
fixed  
none    
none  
   
Send Richard a "before" plot of the single-track test results.
so he can compare
There are no notes attached to this issue.





View Issue Details
391 [Hall D Offline] General minor have not tried 2013-12-17 14:15 2013-12-17 14:15
marki  
marki  
normal  
assigned  
open  
none    
none  
   
Propose a long-term code maintenance scheme
following Paul's discussion at offline meeting.
There are no notes attached to this issue.





View Issue Details
390 [Hall D Offline] General minor have not tried 2013-12-17 14:13 2013-12-17 14:15
marki  
kmoriya  
normal  
assigned  
open  
none    
none  
   
Check Mark's decay chain reporting scheme
to make sure it works
There are no notes attached to this issue.





View Issue Details
359 [JANA] Feature Request minor N/A 2013-06-24 13:33 2013-12-13 17:00
davidl  
davidl  
normal  
resolved  
fixed  
none    
none  
   
Add CopyTo method to JFactory_base
Several instances have occurred where it would be very convenient to be able to copy pointers to externally produced objects into a factory using only the JFactory_base pointer and not the templated JFactory<> pointer. This happens in event sources where several types of objects may be produced but it's not convenient to list al types od data at multiple places in the code.

The CopyTo method should probably take a vector<JObject*> reference and be virtual so that the JFactory<T> method can do the appropriate dynamic_cast before storing it in the factory's _data vector.

It may also be handy to have a templated method of JFactory_base that can take vector<T*>& that it can then use to try and dynamic_cast itself into JFactory<T> and then call the old CopyTo method. (There may be some issues with plugins here so thorough testing will be required.)
Notes
(0000582)
davidl   
2013-12-13 17:00   
Added a CopyTo method to JFactory_base and corresponding one to JFactory. The second suggestion was not implemented. Note that only a compilation test was done here. No testing was done.

This change will appear in JANA 0.7.0





View Issue Details
335 [JANA] Feature Request minor N/A 2013-04-03 10:51 2013-12-13 16:32
davidl  
davidl  
normal  
resolved  
fixed  
none    
none  
   
Add user settable reference to JEventLoop
Sometimes there is need to have an object that is thread specific. For example, a random number generator object in mcsmear. The overhead of maintaining this in a list that one then needs to lock a mutex to access every event is unnecessary. Similarly for creating a factory object and accessing the random number generator through that.

If a user-definable reference were added to JEventLoop, one could set it once and then access it at event time without mutex locking.

This could be done in a simple C-style way where the value is kept as a void* and it is up to the user to cast it back to the appropriate form.

Alternatively, one could use templated methods in JEventLoop to record not only the value, but rtti info so that it can be checked upon retrieval (value would be stored as void* in JEventLoop, but converted by template method before returning). This would be safer, but would add a little more overhead. Presumably not as much as mutex locking though. This second method would also allow for more than one type of reference to be stored, though again, for a slight overhead cost.
Notes
(0000581)
davidl   
2013-12-13 16:32   
This has been committed (with unit test!) and will be in JANA 0.7.0





View Issue Details
387 [JANA] Bug minor have not tried 2013-10-09 07:46 2013-11-18 11:27
davidl  
davidl  
normal  
resolved  
fixed  
none    
none  
   
Event status word reset at wrong time
Here are two e-mail's from Paul:

-----------------------------
Hey David, in BANA I tried setting the status bit in the JEventSource to 1, but by the time I read it in my plugin evnt it has a value of 0. It looks like JEventLoop::OneEvent() calls ClearFactories(), which then clears the status bit. Perhaps the event status bit shouldn't be cleared here? I think it would be better to assume that the JEvent handed to the JEventLoop is already setup properly by the event source. Thanks for taking a look at this.

 - Paul

-----------------------------
Actually, it looks like in JApplication::NextEvent(), you need to manually copy the event status from "myevent" to "event" also. With those two changes, it works for me.

 - Paul
Notes
(0000578)
davidl   
2013-11-18 11:27   
This was committed as svn revision 1395 and will appear in jana 0.6.7. Note that it wasn't clear what Paul meant by "two changes" since ClearFactories call is made before getting the event from the event source and copying the values into the JEventLoop's JEvent.





View Issue Details
388 [JANA] Feature Request minor N/A 2013-11-18 10:41 2013-11-18 11:05
davidl  
davidl  
normal  
resolved  
fixed  
none    
none  
   
Add GetJEventSources method to JApplication
This is requested by Elliott for the farm manager so that the CODAObject-based plugin can get the pointer to the JEventSource reading from ET and tell it to stop if a end run message comes in.

A convenience method to get a pointer to a specific type of JEventSource would be useful too. This would be templated and do the dynamic_cast for the caller. Something like:

T* GetFirstJEventSource()

which would be called like:

JEventSourceEVIO *evio_src = japp->GetFirstJEventSource<JEventSourceEVIO>();
if(evio_src != NULL){
   // ... do something with evio_src ...
}


Notes
(0000577)
davidl   
2013-11-18 11:05   
The requested feature was added in svn version 1394 and will appear in jana 0.6.7





View Issue Details
385 [Hall D Offline] General minor have not tried 2013-09-13 15:13 2013-09-13 15:13
wilevine  
 
normal  
new  
open  
none    
none  
   
BCAL reconstruction: do not consider outer layer when averaging cluster properties
When averaging hit positions and times to form a cluster, we should consider not including hits in the outer layer, due to the shower "curving" issue.
There are no notes attached to this issue.





View Issue Details
384 [Hall D Offline] General minor have not tried 2013-09-13 15:04 2013-09-13 15:04
wilevine  
 
normal  
new  
open  
none    
none  
   
refine code for matching BCAL showers with tracks
Currently the condition for matching BCAL showers and charged tracks is rather arcane (see https://halldweb1.jlab.org/wiki/images/1/18/Bcal_offline_mtg_20130626.pdf [^] ).

One bug is that the matching does not take into account the radial coordinate so you can get matches where the track doesn't even come close to entering the BCAL.

I'm also not sure if delta_phi is calculated correctly.

Furthermore, we should check that the constants used in this condition make sense.
There are no notes attached to this issue.





View Issue Details
383 [Hall D Offline] General minor have not tried 2013-09-13 14:16 2013-09-13 14:16
wilevine  
 
normal  
new  
open  
none    
none  
   
mcsmear BCAL: Simulate ADC timing algorithm
Currently in mcsmear, the times for DBCALHit objects (i.e. ADC hits) are calculated as threshold crossing times, then smeared with 4 ns/sqrt(12) time resolution. In reality, the ADC will not report a threshold crossing timing, but use a more sophisticated algorithm to determine the time. In addition the time resolution may be better than 4 ns/sqrt(12).

The fact that we simulate this as a threshold crossing time means that we need to take an extraneous step of timewalk correction when reconstructing simulated data (that is, we need to do timewalk corrections to the ADC hits in addition to the TDC hits).
There are no notes attached to this issue.





View Issue Details
382 [Hall D Offline] General minor have not tried 2013-09-13 14:05 2013-09-13 14:05
wilevine  
 
normal  
new  
open  
none    
none  
   
mcsmear BCAL: DBCALHit should use GeV instead of ADC units for energy
Currently, mcsmear writes out BCAL hits with energy in pseudo-ADC units rather than in physical units (GeV). This is inconsistent with other detectors and also inconsistent with the new DigiHit scheme, in which the DigiHit objects should use ADC units and the plain Hit objects should use physical units.
There are no notes attached to this issue.





View Issue Details
381 [Hall D Offline] General minor have not tried 2013-09-13 13:50 2013-09-13 13:50
wilevine  
 
normal  
new  
open  
none    
none  
   
mcsmear BCAL: verify BCAL_mevPerPE value
Elton thinks the value of BCAL_mevPerPE is off by a factor of 2:

"We looked at the code in smear_bcal.cc and find the following statement at line 398:
BCAL_mevPerPE = 1.0/( BCAL_PHOTONSPERSIDEPERMEV_INFIBER * BCAL_DEVICEPDE * BCAL_SAMPLING_FRACT );
= 1/(75x 0.21 x 0.095) = 0.67
[As an side, I think this number lacks a factor of 2 reduction in p.e. due to light guide collection, which would bring the factor up to 1.3 MeV/p.e. Is this factor missing?]"
There are no notes attached to this issue.





View Issue Details
380 [Hall D Online] DAQ feature N/A 2013-09-09 17:24 2013-09-09 17:25
wolin  
wolin  
normal  
acknowledged  
open  
none    
none  
   
No
4
0
0
Perform second online data challenge
Second data challenge should include data path starting in ROC's and go all the way to tape, including L3 and event monitoring. Beta version of CODA3 should be used. Use fake data in ROC's. Should use as many crates in the hall as is possible at the time, at least a dozen, and whatever networking exists at the time. Use of the TS and GTP crates probably required. Should push to highest rates possible given the existing hardware, but there is no high-rate goal (that is for the next data challenge).
Notes
(0000576)
wolin   
2013-09-09 17:25   
Dave, Mark D, Sean and probably Justin will be involved.





View Issue Details
360 [Hall D Online] Monitoring feature N/A 2013-06-26 16:54 2013-09-09 17:16
wolin  
wolin  
normal  
resolved  
fixed  
none    
none  
   
No
4
0
0
Monitoring data challenge
Perform monitoring data challenge in late summer. Determine scope and develop goals.
Probable goal is to move data from event builder through monitoring farm to storage on the silo, with creation and archiving of monitoring histograms.
Notes
(0000546)
wolin   
2013-06-26 16:54   
Dave Lawrence and Sean Dobbs will be intimately involved in this.
(0000552)
wolin   
2013-07-18 09:12   
Sean should arrive in late Aug. I created evio2et and et2evio to feed the system and write events to disk. Event transfer to the silo will be either manual or by simple scripts, full system still to be designed by CC. Paul will implement Ganglia for farm monitoring. No farm manager will be used. Dave still working on translation table stuff.
(0000574)
wolin   
2013-09-09 17:16   
Done, currently preparing a document. Very useful, led to many improvements. Next data challenge for late fall, will include data path from ROC to silo, and we'll use CODA3.





View Issue Details
363 [Hall D Online] Monitoring feature N/A 2013-07-18 09:08 2013-09-09 17:15
wolin  
wolin  
normal  
resolved  
fixed  
none    
none  
   
No
1
0
0
Create EVNT plugin
Need EVNT plugin to histogram event sizes, number of words per crate, etc.
Notes
(0000573)
wolin   
2013-09-09 17:15   
Preliminary version developed for online data challenge





View Issue Details
371 [Hall D Online] Computers feature N/A 2013-07-22 10:10 2013-09-09 17:14
wolin  
wolin  
normal  
resolved  
fixed  
none    
none  
   
Yes
1
0
0
Implement preliminary system performance monitoring using Ganglia
We need a performance monitoring system for the data challenge. Paul will implement Ganglia. The CC is thinking of moving to another system eventually
Notes
(0000558)
wolin   
2013-07-22 10:11   
Paul should have time in early Aug.
(0000562)
wolin   
2013-07-26 15:58   
Paul says he started working on this and should have something next week.
(0000572)
wolin   
2013-09-09 17:14   
Paul installed Ganglia on all nodes along with the web server





View Issue Details
378 [JANA] Bug minor always 2013-08-22 08:37 2013-08-27 22:27
davidl  
davidl  
normal  
resolved  
fixed  
none    
none  
   
I/O bound jobs cause excessive CPU use in NextEvent
In JApplication::NextEvent there is a loop that will continually lock/unlock a mutex and check if the application is quitting while waiting for an event to be read in. This means for I/O bound jobs that every thread spends a lot of CPU spinning around that loop while waiting for an event. A short sleep time should be added if the event buffer is empty so that the CPU will be freed up for other things (or at least appear idle) as opposed to showing JANA using it.
Notes
(0000571)
davidl   
2013-08-27 22:27   
This change was applied and will appear in jana 0.6.7





View Issue Details
358 [JANA] Bug minor always 2013-06-18 14:58 2013-08-27 22:14
davidl  
davidl  
low  
resolved  
fixed  
none    
none  
   
Use of typeof not supported when -std=c++11 specified
When compiling with gcc 4.8.0 and the -std=C++11 flag is present, JCalibration.h fails to build due to the use of "typeof". This is apparently a GCC construct and not part of the C++ standard. The webpage linked below suggests using __typeof__ instead and this does seem to work. It's not clear if this is a permanent solution though.

http://gcc.gnu.org/onlinedocs/gcc/Typeof.html [^]
Notes
(0000570)
davidl   
2013-08-27 22:14   
Changed typeof to __typeof__





View Issue Details
377 [Hall D Online] Databases feature N/A 2013-08-09 16:39 2013-08-09 16:42
wolin  
wolin  
normal  
resolved  
fixed  
none    
none  
   
No
1
0
0
Write program to dump PLC archive data
Need a program that provides offline access to PLC archive data
Notes
(0000569)
wolin   
2013-08-09 16:42   
Finished yesterday, python script accepts tag index, database and date range, dumps to file. Currently runs on gluon01 until we move the database. In online scripts directory: DumpPLCArchiveChannel.py





View Issue Details
376 [Hall D Offline] General minor N/A 2013-08-07 10:06 2013-08-07 10:06
davidl  
 
normal  
new  
open  
none    
none  
   
Modify Start Counter sim-=recon to use separate TDC hit
The Start-counter channels have both fADC250 and F1TDC digitization. The current simulation and reconstruction has only one value for the time. In the real data, these will come as separate digitized values that must be linked together in some class. This should be done at the DSCHit level.

The simulation (hdgeant/mcsmear) needs to be modified to provide a high resolution time and low resolution time corresponding to the F1TDC and fADC250 hits respectively.
There are no notes attached to this issue.





View Issue Details
375 [Hall D Offline] General minor N/A 2013-08-07 10:01 2013-08-07 10:01
davidl  
 
normal  
new  
open  
none    
none  
   
Modify TOF sim-recon to use separate TDC hit
The TOF channels have both fADC250 and F1TDC digitization. The current simulation and reconstruction has only one value for the time. In the real data, these will come as separate digitized values that must be linked together in some class. This could be done at the DTOFHit level (omitting the need for DTOFTDCHit) or it could be done at the DTOFPaddleHit level.

The soon-to-be-committed DTOFTDCHit class is not filled by the HDDM event source and the "t" field of DTOFHit is probably too good of resolution for the fADC250.
There are no notes attached to this issue.





View Issue Details
374 [Hall D Online] DAQ feature N/A 2013-07-29 11:13 2013-07-29 11:13
wolin  
wolin  
normal  
acknowledged  
open  
none    
none  
   
No
1
0
0
Create program to start ET systems and configure stations
Need to create ET systems and configure stations in a separate process from the programs that actually use the ET system and attach to the stations. This is especially needed for the L3 and monitoring farm.
Notes
(0000568)
wolin   
2013-07-29 11:13   
I will create a prototype for the upcoming online data challenge.





View Issue Details
331 [Hall D Online] DAQ feature N/A 2013-03-26 17:33 2013-07-26 16:29
wolin  
wolin  
normal  
feedback  
open  
none    
none  
   
No
1
0
0
Install CODA software
Install latest CODA base, libs, packages, etc.
Notes
(0000519)
wolin   
2013-05-16 17:12   
Dave A installed version for high-speed DAQ testing. When he's done I'll figure out the best way to install it. Important question is whether cMsg, evio, etc. should get installed separately or under a single CODA release.
(0000524)
wolin   
2013-06-13 14:23   
Dave Abbott installed CODA3 for DAQ testing. I'm waiting for him to finish basic testing. Until then we don't need our own version of CODA3.
(0000566)
wolin   
2013-07-26 16:29   
Forgot to mention that I installed the evio, et, cMsg and codaObject packages when the gluon file server became operational a few months ago. Not sure if they will remain where they are once the full CODA3 is installed.





View Issue Details
361 [Hall D Online] Monitoring feature N/A 2013-06-26 17:29 2013-07-26 16:00
wolin  
wolin  
normal  
resolved  
fixed  
none    
none  
   
No
1
0
0
evio2et and maybe et2evio
Need program to fill ET system with data from evio file, and maybe another to do the reverse. These will be used in the monitoring data challenge.
Notes
(0000547)
wolin   
2013-06-27 12:20   
And why not et2file and et2xml?
(0000548)
wolin   
2013-06-28 16:16   
Completed evio2et and et2xml. Not sure we need et2evio (same as et2file). Current problem: ET systems bog down for no good reason after a while, probably need to talk to Carl.
(0000551)
wolin   
2013-07-18 09:10   
Created et2evio and evio2cmsg as well. Need to implement run object functionality in evio2et and et2evio.
(0000564)
wolin   
2013-07-26 16:00   
Both evio2et and et2evio are coda objects. I implemented runs in evio2et.





View Issue Details
372 [Hall D Online] Databases minor N/A 2013-07-22 10:30 2013-07-26 15:57
wolin  
wolin  
normal  
feedback  
open  
none    
none  
   
Yes
1
0
0
Upgrade Mantis to the latest version
Minor bugs (default_limit_view), time to upgrade.
Notes
(0000560)
wolin   
2013-07-22 10:30   
Need to hear from Marty/Sherman and Mark I
(0000561)
wolin   
2013-07-26 15:57   
Sherman probably will do the work, need to coordinate with Kelvin and Mark I.





View Issue Details
373 [JANA] Feature Request minor N/A 2013-07-25 15:16 2013-07-25 15:16
davidl  
 
normal  
new  
open  
none    
none  
   
Add ability to call arbitrary routine via janactl
A potentially useful feature would be the ability to send a string to a running JANA process and have it use that to identify a subroutine via dlsym and call it. This would allow someone to write code in one plugin that could then be called remotely via janactl without any specific dispatching code for that routine being written into the janactl callback. It would also allow subroutines compiled into the main executable to be called.

There may be a limitation on the types of arguments that can be passed in via the cMsg. One would also want a message type that can set a global variable to a new value.

While we're at it, we should add getter and setter methods for configuration parameters.
There are no notes attached to this issue.





View Issue Details
314 [Hall D Online] Computers feature N/A 2013-03-21 14:15 2013-07-22 10:14
wolin  
letta  
normal  
assigned  
open  
none    
none  
   
Yes
4
0
0
Install computer and network monitoring system
Choose and install computer monitoring system (Nagios, Ganglia, whatever).
There are no notes attached to this issue.





View Issue Details
353 [Hall D Online] Networking feature N/A 2013-06-07 16:15 2013-07-22 10:13
wolin  
wolin  
normal  
feedback  
open  
none    
none  
   
Yes
8
0
0
Begin developing system to transfer data from RAID to silo staging disks
Investigate strategies to transfer event data from our RAID system to the tape silo staging disks at high speed. Test existing network link and redundant link. Measure throughput.
We will purchase two RAID systems and probably run the event recorder on the active one. Current plans are to take data at 300 MB/s.
Notes
(0000559)
wolin   
2013-07-22 10:13   
CC is working slowly on this. Chris Larrieu is the lead.





View Issue Details
365 [Hall D Online] Databases feature N/A 2013-07-18 10:09 2013-07-22 10:08
wolin  
sergpozd  
normal  
assigned  
open  
none    
none  
   
No
2
0
0
Add new electronics maintenance features to JInventory
Fernando requires a few additions to the JInventory system to make it suitable for long-term electronics maintenance. He will develop a preliminary set of requirements.
Notes
(0000557)
wolin   
2013-07-22 10:08   
Fernando will develop initial set of requirements. After implementation we'll get feedback from technicians, i.e. the people who will use it the most.





View Issue Details
337 [Hall D Online] Magnets feature N/A 2013-04-03 16:24 2013-07-18 09:14
wolin  
wolin  
normal  
feedback  
open  
none    
none  
   
No
2
0
0
Refurbish Danfysik power supply for production running
Refurbish our boards, return Hall C boards, replace other components as needed, get Danfysik techs out here to check everything.
Notes
(0000508)
wolin   
2013-04-03 16:25   
Danfysik techs will be here this summer to work on supplies in other halls. Tim is organizing for them to work on our supply as well.
(0000517)
wolin   
2013-05-16 17:10   
Need to tune for 25H instead of 2H (done for Hall A). Need to program ramp profile to get smooth voltage profile, including no overshoot. Probably need timer to inhibit quench detection when the contacter is first closed for (say) 100 ms.
(0000533)
wolin   
2013-06-14 10:19   
(edited on: 2013-06-14 10:20)
Parts for tuning for 25H on the way from Danfysik. Giles figured out how to program hardware ramp profiles. Dave B will think about doing the same in the PLC.

Note that the hardware has many constraints, including the number of points per profile and the number of profiles stored. The PLC has no constraints. Note that we'll need profiles for ramping up AND down. Further the PLC could generate ramp profiles on the fly based on start/end currents and number of steps, so nothing would need to be stored.

(0000534)
wolin   
2013-06-14 10:21   
Waiting for parts to arrive.
(0000545)
wolin   
2013-06-25 16:59   
Parts here, not sure if we'll install them before or after solenoid test.
(0000553)
wolin   
2013-07-18 09:13   
Since we had innumerable delays in the cryo plant all parts were installed, FPGA programs were updated, etc.
(0000554)
wolin   
2013-07-18 09:14   
We still probably want Danfysik techs to come out and go over the whole power supply.





View Issue Details
324 [Hall D Online] DAQ feature N/A 2013-03-22 10:14 2013-07-18 09:09
wolin  
wolin  
normal  
resolved  
fixed  
none    
none  
   
No
6
0
0
Write out-of-band event handler
Design and implement a cMsg-based mechanism to enter out-of-band events into the event stream (i.e. events not generated by the DAQ system).
Notes
(0000516)
wolin   
2013-05-16 17:08   
Prototype done, based on ipcbank_2et.cc from CLAS online. Needs testing. Uses codaObject framework for communication/control.
(0000550)
wolin   
2013-07-18 09:09   
Implemented this as a run object.





View Issue Details
188 [Hall D Online] Networking feature N/A 2012-01-27 10:14 2013-06-28 16:18
wolin  
wolin  
normal  
acknowledged  
open  
none    
none  
   
No
1
0
0
Install networking in Tagger Building
Install switch in Tagger building, connect via spare JLab dark fibers directly to switch in Rack Room.
Spare dark fibers were or will be included in the bundles between the Tagger building and the nearby service building, between the service building and the Telecom Room, and between the Telecom and Rack Rooms.
Notes
(0000260)
wolin   
2012-01-27 10:15   
Waiting to get Tagger building and for installation of fibers between Telecom and Rack Rooms.
(0000272)
wolin   
2012-01-30 16:47   
Waiting for beneficial occupancy, etc.
(0000537)
wolin   
2013-06-14 10:55   
Waiting for tagger hall to be ready for networking.
(0000544)
wolin   
2013-06-25 16:59   
Tom will get drawing to Jason, who will arrange for fiber run to Tagger hall from service building. Plan is for 12+12 (single+multi-mode) fibers plus one FDU and one switch in the Tagger racks.
(0000549)
wolin   
2013-06-28 16:18   
Jason created the PR, unfortunately for over $11k! I don't understand why it costs so much for 12 single-mode + 24 multi-mode (why so many?) fibers.





View Issue Details
357 [Hall D Online] Monitoring feature N/A 2013-06-14 16:12 2013-06-25 16:57
wolin  
wolin  
normal  
resolved  
fixed  
none    
none  
   
No
1
0
0
Create initial hd_online plugins for monitoring of all detectors
Create preliminary versions of all monitoring plugins. Detector experts will modify these plugins by performing more sophisticated analyses and filling additional histograms.
Initial version will just create occupancy histograms.
Notes
(0000541)
wolin   
2013-06-14 16:13   
I will port the original example plugins I showed at a collaboration meeting a while ago.
(0000542)
wolin   
2013-06-18 11:38   
Almost done creating loadjana.py, loaddana.py and loadroot.py.
(0000543)
wolin   
2013-06-25 16:57   
Plugins for most detectors created and checked in. Modified build system to create plugins. Tested on HDCC (gcc4.8.0) and CUE (gcc4.4.6) systems. Generates basic occupancy histograms, needs improvement by detector builders.





View Issue Details
354 [Hall D Online] Computers feature N/A 2013-06-14 10:12 2013-06-14 15:11
wolin  
wolin  
normal  
resolved  
fixed  
none    
none  
   
No
1
0
0
Add plugin capability to SCONS build system
We need to build plugins as well as binaries and libraries.
Propose plugins directory parallel to libsrc, one subdir of plugins for each plugin, subdir name is plugin name.
Notes
(0000531)
wolin   
2013-06-14 10:13   
With luck this will be a simple modification since plugins work exactly like libraries.
(0000539)
wolin   
2013-06-14 15:11   
Updated build system, scripts and HOWTO to allow for plugins. They are basically handled like libraries but in a different part of the directory hierarchy.





View Issue Details
187 [Hall D Online] Networking feature N/A 2012-01-27 10:10 2013-06-14 10:55
wolin  
wolin  
normal  
feedback  
open  
none    
none  
   
No
4
0
0
Install final networking system in counting house and hall
Install Hall D switch in Rack Room, move existing interim networking from Telecom switch to Rack Room switch, run fibers from Rack Room to platform in the hall, install small switches as needed in the hall.
Notes
(0000259)
wolin   
2012-01-27 10:11   
Will install components as building, rooms, platform, cable trays and racks become available.
(0000271)
wolin   
2012-01-30 16:46   
Waiting for beneficial occupancy, platforms and cable trays to be installed, etc.
(0000493)
wolin   
2013-03-20 18:02   
Counting house near done, will install fiber networking in the hall when cable trays are in, probably this summer.
(0000536)
wolin   
2013-06-14 10:55   
Waiting for cable trays to be installed and the carriage to be ready.





View Issue Details
269 [Hall D Online] DAQ feature N/A 2012-07-10 09:20 2013-06-14 10:54
wolin  
wolin  
normal  
resolved  
fixed  
none    
none  
   
No
1
0
0
Update codaLite to work on Linux-based single board computers
Beni plans to use codaLite for the FDC test. Need to update to use Bryan Moffit's latest VME libraries, latest cMsg, etc.
With luck all that needs to be done is to create a new interrupt service that uses Bryan's libs.
Notes
(0000395)
wolin   
2012-07-10 09:23   
Preliminary version of new interrupt service created with Bryan's help, needs to be tested. Beni and I set up wilma with a 32-bit operating system (required for Bryan's libs), put it on the 37 subnet along with the SBC's, and installed and tested the latest cMsg.

Now we're ready to test codaLite. It is an old version that Beni has been using for a long time. Decided not to update at this time.
(0000399)
wolin   
2012-07-16 11:29   
(edited on: 2012-07-16 11:32)
Linux ROC polling interrupt service for old TI's appears to work, requires gef kernel module and Bryan's linuxvme libraries. This interruptService polls the old TI to check if a trigger is pending. wilma back on 88 subnet. Will create polling interrupt service for new TI's when Beni gets them back. At some point should skip the polling and use interrupt service routine instead. Bryan says this is all implemented in his libraries so I don't have to redo what I did for VxWorks.

(0000400)
wolin   
2012-07-16 11:33   
Beni is happy now. Will have to revisit this when he wants to run with new TI's.
(0000535)
wolin   
2013-06-14 10:54   
Beni got an interrupt service working on Linux with the new TI board. It is somewhat tricky and complicated. I'll try to generalize his code when an application arises, until then Beni is the only one using it.





View Issue Details
336 [Hall D Online] Controls feature N/A 2013-04-03 16:20 2013-06-14 10:16
wolin  
wolin  
normal  
feedback  
open  
none    
none  
   
No
1
0
0
Install ControlNet cable between solenoid PLC and cryo PLC
Install cable if we decide this is the way to go.
Notes
(0000518)
wolin   
2013-05-16 17:11   
Jonathan more interested now that the magnet can quench due to adverse conditions in the cryo plant.
(0000532)
wolin   
2013-06-14 10:16   
Need confirmation from Cryo group and agreement on who controls what.





View Issue Details
227 [Hall D Online] Infrastructure minor N/A 2012-04-04 15:21 2013-06-13 14:31
wolin  
wolin  
normal  
resolved  
fixed  
none    
none  
   
No
1
0
0
Determine quality of power in Hall D complex
Ask Jonathan to install power monitor in Hall D complex so we can get a preliminary idea on the frequency of dropouts.
Notes
(0000335)
wolin   
2012-04-04 15:22   
Contacted Jonathan, who is out of town this week. He says he will do this.
(0000364)
wolin   
2012-04-24 15:07   
Jonathan says he installed the power monitor and is collecting data.
(0000366)
wolin   
2012-05-08 11:38   
So far no dropouts or anomolous behaviour, just the usual minor voltage swings. We're aiming for at least one full month of data collection.
(0000530)
wolin   
2013-06-13 14:31   
No problems after many months of solenoid testing, I guess this isn't a problem. We never got a working line monitor on the power lines.





View Issue Details
283 [Hall D Online] Databases feature N/A 2012-09-11 14:02 2013-06-13 14:29
wolin  
sergpozd  
normal  
assigned  
open  
none    
none  
   
No
4
0
0
Generate graphic web pages using JInventory database information
Draw racks, crates, modules, etc. on web pages. Allow drill-down from rooms to racks to crates to modules. Clicking on items should bring up JInv entry for that item.
There are no notes attached to this issue.





View Issue Details
338 [Hall D Online] DAQ feature N/A 2013-04-08 09:17 2013-06-13 14:24
wolin  
 
normal  
new  
open  
none    
none  
   
No
1
0
0
Decide on compressed data format
Get agreement from DAQ group concerning a compressed data format, same as was done for the uncompressed format.
Dave A says they need more time working with the DAQ system before they will be ready to do this. In the mean time we'll use Elliott's provisional format.
Notes
(0000525)
wolin   
2013-06-13 14:24   
Waiting for CODA3 to settle down a bit. Dave Abbott said they are not ready to decide this yet.





View Issue Details
352 [Hall D Online] Databases feature N/A 2013-06-07 14:59 2013-06-13 08:48
wolin  
sergpozd  
normal  
acknowledged  
open  
none    
none  
   
No
12
0
0
Upgrade JInventory database system
Develop conventions and update/upgrade database and php code to implement them. Create and implement procedure to add electronics information to database from Fernando's spreadsheets. Purchase and install barcode labels on as much equipment as possible and update database appropriately. Add power hierarchy information to database.
Dave L's SQLite database may be useful for getting at electronics information. Some items already have barcode labels. Use label printer to make large labels for selected equipment. Work with Mark Stevens on power hierarchy.
Notes
(0000523)
wolin   
2013-06-13 08:48   
We've decided to add front-end boards using the CLAS convention of a fixed five-layer housing hierarchy. Serguei P will port the CLAS gui's that implement this into the system and modify them to use the housing hierarchy instead of location tables. These gui's restrict the fields shown to those relevant for front-end electronics boards. The full item gui will be used for other items.





View Issue Details
186 [Hall D Online] General feature N/A 2012-01-23 12:21 2013-06-07 14:47
wolin  
wolin  
high  
resolved  
fixed  
none    
none  
   
No
5
0
0
Develop scons-based build system
Develop generic scons-based build system for online software and appropriate online directory structure.
Review and steal best ideas from BMS, CLAS12 scons system, DAQ scons system, etc.
Notes
(0000253)
wolin   
2012-01-23 14:56   
Maurizio will present the CLAS12 scons system at the next Hall D Online meeting.
(0000504)
wolin   
2013-03-21 14:11   
Maurizio's system is the best place to start with. He would welcome joint development of his system with us.
(0000514)
wolin   
2013-05-14 11:34   
Basis build and install scripts complete, package directory structure defined. Still need to create release scripts for use by hdsys account.
(0000515)
wolin   
2013-05-16 17:07   
Release script done, just need script from Dave L that generates setenv statements and creates setenv_xxx script.
(0000522)
wolin   
2013-06-07 14:47   
System working and checked into svn. Scripts for build and release created.





View Issue Details
296 [Hall D Offline] General minor have not tried 2012-12-03 11:17 2013-05-21 10:26
marki  
staylor  
normal  
resolved  
fixed  
none    
none  
   
fix the cdc geometry
problem reported at the 11/28/12 offline meeting
Notes
(0000521)
staylor   
2013-05-21 10:25   
updated last fall





View Issue Details
347 [Hall D Offline] General minor have not tried 2013-05-16 12:54 2013-05-16 12:54
marki  
 
normal  
new  
open  
none    
none  
   
discuss the curling track problem
Paul and Simon
There are no notes attached to this issue.





View Issue Details
346 [Hall D Offline] General minor have not tried 2013-05-16 12:53 2013-05-16 12:53
marki  
 
normal  
new  
open  
none    
none  
   
summarize the detector naming convention in a document
based on translation tables
There are no notes attached to this issue.





View Issue Details
345 [Hall D Offline] General minor have not tried 2013-05-16 12:53 2013-05-16 12:53
marki  
 
normal  
new  
open  
none    
none  
   
add aluminum manifold for FDC pre-amp cooling in simulation
not in there now
There are no notes attached to this issue.





View Issue Details
344 [Hall D Offline] General minor have not tried 2013-05-16 12:52 2013-05-16 12:52
marki  
 
normal  
new  
open  
none    
none  
   
add material in front of BCAL in simulation
light pulsing system material
There are no notes attached to this issue.





View Issue Details
343 [Hall D Offline] General minor have not tried 2013-05-16 12:52 2013-05-16 12:52
marki  
 
normal  
new  
open  
none    
none  
   
fix TOF geometry
include half-width counters
There are no notes attached to this issue.





View Issue Details
342 [Hall D Offline] General minor have not tried 2013-05-16 12:51 2013-05-16 12:51
marki  
 
normal  
new  
open  
none    
none  
   
revisit data challenge 1
look for issues to solve for dc2
There are no notes attached to this issue.





View Issue Details
341 [Hall D Online] Computers feature N/A 2013-05-14 11:30 2013-05-14 11:31
wolin  
wolin  
normal  
resolved  
fixed  
none    
none  
   
No
1
0
0
Set up gluex accounts on HDCC
Set up hdsys, hdops and user accounts and home directories
There are no notes attached to this issue.





View Issue Details
330 [Hall D Online] Computers feature N/A 2013-03-26 17:32 2013-05-14 11:22
wolin  
wolin  
normal  
resolved  
fixed  
none    
none  
   
No
1
0
0
Set up gluex directories on HDCC
Set up source, distribution, etc. areas on file server.
There are no notes attached to this issue.





View Issue Details
328 [Hall D Offline] General minor N/A 2013-03-26 16:02 2013-05-05 16:09
davidl  
wilevine  
normal  
resolved  
fixed  
none    
none  
   
sim-recon soft-wired for KLOE
The sim-recon code has several places where special configuration parameters like "BCALRECON:USE_KLOE" are used to specify use of the DBCALShower objects produced using the KLOE algorithm. This duplicates the same functionality already in JANA available via the DEFTAG:XXX configuration parameter. These should be removed so that the default tag can be overridden via the built-in JANA mechanism.

In addition, Beni has recently reported that the KLOE algorithm seems to be doing much worse than the IU algorithm. When this issue is fixed, the IU algorithm should probably be made the default.
Notes
(0000512)
marki   
2013-05-05 16:09   
Will reported on this at the May 1, 2013 offline meeting.





View Issue Details
33 [Hall D Offline] General minor have not tried 2010-11-04 13:46 2013-05-01 10:57
marki  
romanov  
normal  
resolved  
fixed  
none    
none  
   
Try importing GlueX geometry into EVE
Beni and Dmitry should get together and try to import our HDDS-based geometry into EVE.
Notes
(0000275)
marki   
2012-01-31 14:26   
Reminder sent to: zihlmann
Is there a possibility of reviving this? Perhaps we should discuss it.
(0000511)
marki   
2013-05-01 10:57   
Dmitry has demonstrated that the import can be done.





View Issue Details
133 [Hall D Offline] General minor have not tried 2011-09-21 12:59 2013-05-01 10:54
marki  
zihlmann  
normal  
resolved  
fixed  
none    
none  
   
Beni will present HERMES particle ID scheme
for our education
Notes
(0000510)
marki   
2013-05-01 10:54   
Beni distributed document describing Hermes scheme





View Issue Details
301 [Hall D Offline] General minor have not tried 2013-02-09 17:47 2013-05-01 10:47
marki  
marki  
normal  
resolved  
fixed  
none    
none  
   
b1pi/single-track results notification
need to implement a notification scheme
Notes
(0000509)
marki   
2013-05-01 10:47   
done several weeks ago using simple_email_list





View Issue Details
228 [Hall D Online] Magnets feature N/A 2012-04-04 15:31 2013-04-03 16:22
wolin  
stevensm  
normal  
resolved  
fixed  
none    
none  
   
No
6
0
0
Repair/upgrade Danfysik power supply for solenoid test
Implement the many fixes/improvements we determined need to be done for the Fall test. These include instrumenting for the SOE module, new ground fault detector scheme, wiring of UPS power to contactor and control boards, replace Hall C control board with our repaired board, installation of key lock box, update checkout/commissioning procedures, backup of internal board program, etc.
Notes
(0000369)
wolin   
2012-05-09 11:37   
Current plans are to have Bill Merz's group do routine maintenance, upgrades and repairs to the power supply in cooperation with Hall D. They can upgrade boards, upgrade firmware, etc. No need to bring in Danfysik techs.
(0000371)
stevensm   
2012-05-22 11:54   
I have marked up a "Main Schematic" print with my proposed SOE (Sequence Of Events) relays, and the change in the ground fault circuit. I am waiting for Giles Brown to go over the print with me to approve of the modifications.
(0000505)
wolin   
2013-03-25 14:48   
Mark and folks from Accel are repairing boards, replacing power transistors, etc. Current tests look good.
(0000507)
wolin   
2013-04-03 16:22   
The supply is ready to go. Many transistors replaced, some boards replaced (borrowed from Hall C).





View Issue Details
334 [JANA] Feature Request minor N/A 2013-03-27 18:56 2013-03-27 19:25
davidl  
davidl  
normal  
resolved  
fixed  
none    
none  
   
Add IsAssociated method to JObject
Paul M. requests the following method be added to JObject:


inline bool JObject::IsAssociated(const JObject* locObject) const {return (associated.find(locObject) != associated.end());}
Notes
(0000506)
davidl   
2013-03-27 19:25   
This was added in revision 1273 and will appear in the next tagged version of JANA (either 0.6.6p2 or 0.6.7)





View Issue Details
329 [Hall D Offline] General minor always 2013-03-26 16:07 2013-03-26 16:07
davidl  
 
normal  
new  
open  
none    
none  
   
Add thread-specific random number generators to mcsmear
The mcsmear program currently uses a single random number generator for all threads. While it
writes the random number seeds to the output HDDM file at the beginning of the event, the sequence will not be reproducible if more than one thread is used. To fix this problem, dedicated random number generators need to be instantiated for each thread.
There are no notes attached to this issue.





View Issue Details
310 [Hall D Online] Controls feature N/A 2013-03-21 13:51 2013-03-22 13:25
wolin  
debutler  
normal  
acknowledged  
open  
none    
none  
   
No
1
0
0
Learn about PXI programming for solenoid controls
Yi created the PXI control program in Labview. Someone else needs to learn about it.
There are no notes attached to this issue.





View Issue Details
139 [Hall D Online] Controls feature N/A 2011-09-21 16:18 2013-03-22 13:25
wolin  
debutler  
normal  
acknowledged  
open  
none    
none  
   
No
8
0
0
Determine requirements for target control system, design and implement
The cryotarget will be tested in Summer 2013. We will need the complete control system by then.
Chris Keith is designing the cryo system. He will tell us what instrumention will be used.
Notes
(0000166)
wolin   
2011-10-12 14:51   
Chris Keith is designing the target system.
(0000167)
wolin   
2011-10-12 14:51   
Waiting to hear from Chris Keith...
(0000492)
wolin   
2013-03-20 18:01   
Mostly designed by now, we can communicate with the Lakeshore controller via 495NBX module, details still to be worked out.





View Issue Details
318 [Hall D Online] Controls feature N/A 2013-03-21 17:29 2013-03-22 13:24
wolin  
debutler  
normal  
acknowledged  
open  
none    
none  
   
No
4
0
0
Determine requirements for FCAL dark room system, design and implement
Purchase PLC equipment, implement control algorithm. The dark room should be ready before summer 2013.
There are no notes attached to this issue.





View Issue Details
320 [Hall D Online] Controls feature N/A 2013-03-21 17:31 2013-03-22 13:23
wolin  
debutler  
normal  
acknowledged  
open  
none    
none  
   
No
4
0
0
Determine requirements for gas system, design and implement
Purchase PLC equipment, implement control algorithm.
There are no notes attached to this issue.





View Issue Details
321 [Hall D Online] Controls feature N/A 2013-03-21 17:32 2013-03-22 13:23
wolin  
debutler  
normal  
acknowledged  
open  
none    
none  
   
No
4
0
0
Determine requirements for BCAL temperature monitoring system, design and implement
Purchase PLC equipment, implement control algorithm.
There are no notes attached to this issue.





View Issue Details
319 [Hall D Online] Controls feature N/A 2013-03-21 17:30 2013-03-22 13:23
wolin  
debutler  
normal  
acknowledged  
open  
none    
none  
   
No
2
0
0
Determine requirements for BCAL cooling system, design and implement
Purchase PLC equipment, implement control algorithm.
There are no notes attached to this issue.





View Issue Details
322 [Hall D Online] Controls feature N/A 2013-03-21 17:33 2013-03-22 13:23
wolin  
debutler  
normal  
acknowledged  
open  
none    
none  
   
No
2
0
0
Determine requirements for FCAL cooling system, design and implement
Purchase PLC equipment, implement control algorithm.
There are no notes attached to this issue.





View Issue Details
327 [Hall D Online] Monitoring feature N/A 2013-03-22 11:22 2013-03-22 11:22
wolin  
 
normal  
new  
open  
none    
none  
   
No
6
0
0
Develop hardware monitoring facility
Develop facility to monitor various types of hardware, including crates, ROC's, etc. It is possible that the sysadmin computer monitoring facility can do all of this, but I suspect not, so an additional facility must be developed. This facility must interoperate with DAQ run control.
There are no notes attached to this issue.





View Issue Details
326 [Hall D Online] Monitoring feature N/A 2013-03-22 11:17 2013-03-22 11:17
wolin  
 
normal  
new  
open  
none    
none  
   
No
4
0
0
Develop process monitoring system
Develop system to monitor health of computer processes running on counting house computers. A separate facility will monitor system processes, this facility needs to monitor user processes.
There are no notes attached to this issue.





View Issue Details
323 [Hall D Online] Databases feature N/A 2013-03-22 10:12 2013-03-22 10:12
wolin  
 
normal  
new  
open  
none    
none  
   
No
6
0
0
Write run bookkeeping scripts
Write scripts that enter critical run-related information at various stages during the run. Leading strategy is to enter everything into the conditions database, but a backup strategy needs to be developed as well. The latter likely should be file-based, and multiple backups might be a good idea.
There are no notes attached to this issue.





View Issue Details
231 [Hall D Online] Computers feature N/A 2012-04-04 15:37 2013-03-21 14:01
wolin  
ballard  
normal  
resolved  
fixed  
none    
none  
   
No
1
0
0
Install computer in controls racks
Install windows computer in controls racks, install Allen-Bradley software as needed.
Notes
(0000393)
ballard   
2012-06-29 13:59   
(edited on: 2012-06-29 13:59)
Purchased quad-core developer PC for hosting HMI/RSLinx Enterprise servers.
(0000502)
wolin   
2013-03-21 14:01   
Computer installed, may want to put a better one in there some day.





View Issue Details
72 [Hall D Online] Controls feature N/A 2011-04-21 09:28 2013-03-21 13:59
hovanes  
 
normal  
new  
open  
none    
none  
   
Yes
8
0
0
Top-level Experiment Control System supervisor and GUI
Design a top-level Experiment Control System (ECS) supervisor that will supervise CODA run control as well as many other lower-level controls components/supervisors (monitoring/L3 farm, trigger system, HV, LV, etc). Design includes the ECS state machine and its associated operator GUI. Develop prototype and test in a realistic DAQ and controls environment. Deliver report that summarizes results and includes an outline of steps needed to create and commission a production version with estimates of required time and manpower.




AFECS is expected to be the top level layer for configuring and running the Hall D experiment. In this task we need to design an interface between AFECS and Hall D slow controls system (once it is defined), and create some prototypes. We will need to:
* Determine what needs to be communicated between Hall D slow controls system and other AFECS components , like Run Control, Alarms, Monitoring.
* Define how control system interacts with AFECS
* Design slow controls object(s) to communicate with AFECS
* Design an interface between AFECS and Hall D controls screen management system.
Notes
(0000194)
wolin   
2011-10-31 17:21   
Since Hall D controls will use the CSS package to develop operator gui's, it makes sense to use CSS for the ECS gui as well. CSS is composed of Eclipse plugins that can communicate via Channel Access. It is not clear how difficult it is to extend the CSS base classes to communicate via cMsg, required for the gui to talk to AFECS supervisor components. Vardan is looking into this.
(0000498)
wolin   
2013-03-21 13:59   
Not sure who will work on this.





View Issue Details
247 [Hall D Online] Computers feature N/A 2012-04-20 09:04 2013-03-20 18:05
wolin  
 
normal  
new  
open  
none    
none  
   
No
2
0
0
Learn how to query the status of the counting house and other UPS's
The control system needs to know if UPS's are running on line or battery power, and how much battery power is left. Figure out how to obtain this information for our existing UPS's and how this information might be accessed by the PLC and EPICS.
Notes
(0000373)
wolin   
2012-05-30 17:09   
CNU summer student Rob Ryan will be looking into this. Mike Johnson has an EPICS way of doing this, Vardan can do this in AFECS. Current thinking is that the AFECS mechanism is the way to go.
(0000389)
wolin   
2012-06-25 15:07   
From Mark Stevens:

I recall that you are perhaps having a student look at the software/internet monitoring of the Libert UPS in the counting room. Are you aware that the RDC has the capability of communicating too, with optional equipment? And are you aware that the RDC may have a built-in monitoring system that self-protects and sends out alarms for overcurrent, and that it is factory-programmed for single-phase 20 Amp circuits? Will we need to reprogram this system to avoid erroneous faults, because we are installing larger circuits? I read about all of this in the Liebert RDC manual, while I was looking for information about circuit breakers. Perhaps your student could check this out.
(0000495)
wolin   
2013-03-20 18:04   
Rob is gone, no progress...





View Issue Details
151 [Hall D Online] EPICS feature N/A 2011-10-21 06:42 2013-03-20 17:59
wolin  
nerses  
normal  
resolved  
fixed  
none    
none  
   
No
4
0
0
Develop EPICS driver for JLab discriminator board
Develop driver to allow EPICS to communicate with the JLab discriminator board.
Joint effort with Hall B since we both will be using this board.
Notes
(0000183)
nerses   
2011-10-21 09:22   
The project is started
(0000227)
wolin   
2011-12-07 11:46   
Reminder sent to: nerses
Kindly update this task as you make progress, thanks.





View Issue Details
238 [Hall D Online] EPICS feature N/A 2012-04-04 16:04 2013-03-20 17:59
wolin  
hovanes  
normal  
resolved  
open  
none    
none  
   
No
1
0
0
Develop EPICS PV naming scheme
Develop EPICS PV naming scheme, present to Controls group.
There are no notes attached to this issue.





View Issue Details
224 [Hall D Online] EPICS feature N/A 2012-04-04 15:15 2013-03-20 17:59
wolin  
hovanes  
normal  
resolved  
open  
none    
none  
   
No
4
0
0
Develop EPICS driver for Tagger microscope
Develop driver and IOC to control microscope boards.
Richard's boards currently communicate at the Ethernet Transport level, and do not understand IP. Initial indications are that this can be handled by our network equipment.
There are no notes attached to this issue.





View Issue Details
223 [Hall D Online] EPICS feature N/A 2012-04-04 15:12 2013-03-20 17:59
wolin  
hovanes  
normal  
resolved  
open  
none    
none  
   
No
4
0
0
Develop EPICS driver for FCAL bases
Develop driver and IOC to control FCAL bases.
There are no notes attached to this issue.





View Issue Details
222 [Hall D Online] EPICS feature N/A 2012-04-04 15:11 2013-03-20 17:59
wolin  
hovanes  
normal  
resolved  
open  
none    
none  
   
No
4
0
0
Develop EPICS driver for Wiener LV mainframes
Develop driver and IOC to control Wiener LV system
There are no notes attached to this issue.





View Issue Details
221 [Hall D Online] EPICS feature N/A 2012-04-04 15:10 2013-03-20 17:59
wolin  
hovanes  
normal  
resolved  
open  
none    
none  
   
No
4
0
0
Develop EPICS driver for CAEN HV mainframes
Develop driver and IOC to control CAEN HV
There are no notes attached to this issue.





View Issue Details
233 [Hall D Online] Databases feature N/A 2012-04-04 15:41 2013-03-19 13:10
wolin  
stevensm  
normal  
resolved  
fixed  
none    
none  
   
No
2
0
0
Insert manuals into Midas elog system
Need to get all our manuals into the Midas manual database system.
Various people will work on this, Mark is first.
Notes
(0000340)
wolin   
2012-04-04 15:42   
Mark, once you are finished reassign this to the next person, whoever that might be.
(0000390)
stevensm   
2012-06-28 16:32   
Elliott, I have responded to this for all of the people who work for/with me. I have no power over anyone else. Whom should I transfer this to?
(0000491)
wolin   
2013-03-19 13:10   
Mark et al. put a great many manuals in, this of course is an ongoing task...





View Issue Details
257 [Hall D Online] Magnets feature N/A 2012-04-24 15:04 2013-03-19 13:09
wolin  
stevensm  
normal  
resolved  
fixed  
none    
none  
   
No
1
0
0
Develop test procedures for solenoid Danfysik supply
Procedure should especially test the slow dump, as we think never performed a slow dump during the coil tests.
Sarin Philip said he'd help to develop this.
Notes
(0000490)
wolin   
2013-03-19 13:09   
Mark and others created a maintenance OSP for the Danfysik supply.





View Issue Details
235 [Hall D Online] Controls feature N/A 2012-04-04 15:45 2013-03-19 13:09
wolin  
stevensm  
normal  
resolved  
fixed  
none    
none  
   
No
2
0
0
Design and implement SOE monitoring of power supply trip chains
Design circuits, procure parts and implement SOE monitoring of as many components of the Danfysik trip chains as is feasible.
Notes
(0000489)
wolin   
2013-03-19 13:09   
Completed recently, I think it was fully tested, not sure.





View Issue Details
230 [Hall D Online] Controls major N/A 2012-04-04 15:36 2013-03-19 13:08
wolin  
spiegel  
normal  
resolved  
fixed  
none    
none  
   
No
6
0
0
Design and build new breakout box for solenoid controls testing
Design and construct four-coil breakout box for the solenoid controls system.
Notes
(0000488)
wolin   
2013-03-19 13:08   
Completed a while ago.





View Issue Details
225 [Hall D Online] Controls feature N/A 2012-04-04 15:18 2013-03-19 13:08
wolin  
ballard  
normal  
resolved  
fixed  
none    
none  
   
No
8
0
0
Rewrite PLC program for full solenoid test
Modify, upgrade or rewrite PLC code as appropriate for full solenoid test.
Use ladders, function block or structured text as appropriate. Maximize modularity and reuse via liberal use of add-on instructions.
Notes
(0000487)
wolin   
2013-03-19 13:08   
Josh and Dave B finished doing this.





View Issue Details
163 [Hall D Online] General feature N/A 2011-11-04 09:32 2013-03-19 13:07
wolin  
hovanes  
normal  
resolved  
fixed  
none    
none  
   
No
4
0
0
Outfit Counting House
We will get the Counting House in the next few weeks. We need to begin design of the control room and order furniture. Installation should be performed Summer 2012.
Facilities has expertise outfitting work areas. We'll need control consoles and chairs, tables and chair for the work/conference area, storage shelves and cabinets, refrigerator and microwave.
Notes
(0000196)
wolin   
2011-11-04 09:34   
I will soon contact Facilities and work with Dianne to figure out how to proceed on this. We'll need some interim furniture before we purchase the good stuff.
(0000249)
wolin   
2012-01-23 10:00   
We may get some surplus furniture from the Test Lab when the technical staff moves to the new TEDF building. I told Dave Fazenbaker and Lamont we need all kinds of stuff.
(0000486)
wolin   
2013-03-19 13:07   
Almost everything installed or ordered, not much left at this point...





View Issue Details
179 [Hall D Online] Controls feature N/A 2011-12-19 13:31 2013-03-19 13:06
wolin  
hovanes  
normal  
resolved  
open  
none    
none  
   
No
1
0
0
Write planning document for archiving controls information
Document covers archiving EPICS information
There are no notes attached to this issue.





View Issue Details
178 [Hall D Online] General feature N/A 2011-12-19 13:29 2013-03-19 13:06
wolin  
wolin  
normal  
resolved  
open  
none    
none  
   
No
1
0
0
Write planning document for archiving run information
Document covers archiving all information other than DAQ configurations and EPICS archive data
There are no notes attached to this issue.





View Issue Details
171 [Hall D Online] Monitoring feature N/A 2011-11-21 09:45 2013-03-19 13:06
wolin  
davidl  
normal  
resolved  
open  
none    
none  
   
No
1
0
0
Write planning document for event display
Write planning document based on previous work by Andrew Blackburn, Dave Heddle, Dave Lawrence and others. Include online requirements, overlap with offline requirements, ET and file event source, interoperation with JANA and JANA objects.
Notes
(0000214)
wolin   
2011-11-21 09:45   
Time to get going on 12GeV planning documents...
(0000215)
wolin   
2011-11-21 09:45   
Low hanging fruit...





View Issue Details
253 [Hall D Online] Controls feature N/A 2012-04-24 14:21 2013-03-19 13:05
wolin  
creel  
normal  
resolved  
fixed  
none    
none  
   
No
1
0
0
Install ControlNet cable between Hall D and Cryo Service building
Communication between the Hall D and Cryo PLC systems will be done via ControlNet. A cable needs to be installed between the two buildings.
Jonathan said he has the cable and connectors. He said his techs will actually install the cable.
Notes
(0000368)
wolin   
2012-05-09 11:32   
Best strategy seems to be to run the line to the PLC chassis in Rack Room.
(0000485)
wolin   
2013-03-19 13:05   
We decided this isn't needed. We may install on at a later date.





View Issue Details
153 [Hall D Online] Controls feature N/A 2011-10-24 11:37 2013-03-19 12:28
wolin  
hovanes  
normal  
resolved  
fixed  
none    
none  
   
No
4
0
0
Survey EPICS archivers and choose one
Survey existing EPICS archiver packages, test and make recommendation.
(not that important any more...see note below, EJW) An important consideration is integration/interoperation with the CCDB (Calibration and Conditions Database) to meet the needs of the offline analysis effort.
Notes
(0000203)
wolin   
2011-11-15 11:09   
Hovanes is focusing in on the SNS archvier (RDB/Beauty) and the JLab Accelerator archiver (MyA).
(0000204)
wolin   
2011-11-15 11:11   
Interoperation with the CCDB seems less important now. The current thinking is to sweep data out of the archiver and process it before making an entry into the CCDB. Offline processes will then just access the CCDB and never directly access the archiver database.
(0000208)
hovanes   
2011-11-16 09:04   
Work in progress
(0000484)
wolin   
2013-03-19 12:28   
We will use the JLab Accelerator div's MyA. We will purchase some computers for them to run our archiver, Matt will buy them when needed (but before Apr-2014!).





View Issue Details
234 [Hall D Online] Controls feature N/A 2012-04-04 15:43 2013-03-19 12:26
wolin  
ballard  
normal  
resolved  
fixed  
none    
none  
   
No
1
0
0
Procure and test 490NBX module
Purchase, install and test 490NBX module with some system that accepts Ethernet control.
Notes
(0000344)
ballard   
2012-04-04 16:00   
We have the 490NBX and it has been issued hostname halld_490nbx1.jlab.org and IP address 129.57.36.37. It is installed in the PLC development area in EEL 126 but has not yet been configured.
(0000483)
wolin   
2013-03-19 12:26   
Purchased a while ago, works perfectly with the Lakeshore controller.





View Issue Details
194 [Hall D Online] Infrastructure feature N/A 2012-02-01 09:08 2013-03-19 12:26
wolin  
hovanes  
normal  
resolved  
fixed  
none    
none  
   
No
1
0
0
Install UPS in the rack room
Install UPS, wire it up, install RDC.
Notes
(0000328)
hovanes   
2012-04-03 09:07   
This task is actually managed by Giles Brown. Hovanes Egiyan is the POC.
(0000336)
wolin   
2012-04-04 15:25   
Need to determine whether UPS generates too much noise to allow us to run UPS power to the hall.
(0000347)
wolin   
2012-04-04 17:46   
Harris running power from utility room this week or early next week, Tom will move UPS in before then.
(0000482)
wolin   
2013-03-19 12:26   
This was completed a while ago.





View Issue Details
236 [Hall D Online] Controls feature N/A 2012-04-04 15:48 2013-03-19 12:25
wolin  
wolin  
normal  
resolved  
fixed  
none    
none  
   
No
1
0
0
Modify Danfysik quench detector for use with solenoid
Resistors on the quench detector must be modified to give acceptable adjustment range, taking into account current-limiting resistors in the coil tap circuits.
Notes
(0000342)
wolin   
2012-04-04 15:49   
Need signed drawings for coil tap circuits. Best would be for all four QD channels to be identical, which might require additions to the current-limiting resistor boxes. Or we might eliminate the redundant QD channel to avoid this.
(0000424)
wolin   
2012-08-08 14:48   
Alexandre Elkiaer though my first redesign had too much input impedance, he wants to minimize this based on the maximum quench voltage. George will try to get this to me. With this I can calculate the resistor and pot values needed in the input circuits.
(0000433)
wolin   
2012-08-20 10:26   
Sent two proposals to Alexander with different input impedances, not sure which is best. He's on vacation for a while...
(0000439)
wolin   
2012-09-11 13:57   
Alexander suggested going with the lower impedance version, and described how to set the cable-detect resistors.
(0000481)
wolin   
2013-03-19 12:25   
Front-end resistors modified for full solenoid, POT2 not changed. Three channels tuned for different coil comparisons. Cable detect resistors NOT modified so this feature is currently not working. Not clear if we need it since there is so much redundancy in the overall QD system.





View Issue Details
189 [Hall D Online] Networking feature N/A 2012-01-27 10:17 2013-03-19 12:00
wolin  
wolin  
normal  
resolved  
fixed  
none    
none  
   
No
1
0
0
Install fibers between Telecom and Rack Rooms
Main connection between our switch and the JLab backbone networking system.
Originally was included in the 12GEV civil contract but got cut out. We may have to pay for it. Latest quote from Shields was over $8k.
Notes
(0000262)
hovanes   
2012-01-27 10:49   
The online schedule is being updated to allocate enough money for fiber installations in the counting house itself. The activity line number is going to be 1531048.
(0000263)
wolin   
2012-01-27 11:17   
From Hovanes:

We are adding 10K to the schedule, activity line 1531048. Hopefully, it gets approved, but we should wait for approval. We are not installing the switch right now anyway. Please, make sure it gets charged to "12CDAQ.12GEV" POA, with "Activity line 1531048" in the Notes for PR.
(0000306)
wolin   
2012-02-08 15:42   
(edited on: 2012-02-08 16:14)
Dianne says no problem with getting fibers, few day installation. Brian says prior to installing the fibers we just need the Telecom rack installed, no need for power, and he prefers that the switch go in after the fibers.

(0000363)
wolin   
2012-04-24 15:07   
Waiting for UPS to be installed, then for power distribution installation under the floor. Then we can move the racks in and install the fibers.
(0000381)
wolin   
2012-06-05 11:41   
Waiting for counting house rack room power to be ready and racks installed, meanwhile we get temporary networking from the Telecom room switch.
(0000423)
wolin   
2012-08-08 14:46   
Power and racks in Rack Room should be complete by early Sep. I contacted Bryan Hess and asked that he plan to install the fibers (and our network switch) in mid-Sep, assuming all goes well with the electrical and mechanical work.
(0000438)
wolin   
2012-09-11 13:56   
Jason got a quote from Shields, $8k!!! Why so much?
(0000480)
wolin   
2013-03-19 12:00   
This was completed a while ago.




View Issue Details
152 [Hall D Online] Electronic logbook feature N/A 2011-10-21 09:17 2013-03-19 11:59
wolin  
wolin  
normal  
resolved  
fixed  
none