OWG Meeting 7-Sep-2016
Location and Time
Room: CC F326
You can connect using BlueJeans Video conferencing (ID: 120 390 084). (Click "Expand" to the right for details -->):
(if problems, call phone in conference room: 757-269-6460)
- To join via Polycom room system go to the IP Address: 220.127.116.11 (bjn.vc) and enter the meeting ID: 120390084.
- To join via a Web Browser, go to the page https://bluejeans.com/120390084.
- To join via phone, use one of the following numbers and the Conference ID: 120390084
- US or Canada: +1 408 740 7256 or
- US or Canada: +1 888 240 2560
- More information on connecting to bluejeans is available.
- New equipment installation
- New nodes gluon112-gluon119, gluonraid3
- IB installed/configured (Dave A.)
- 10Gbit ethernet on gluon113 and gluon115 (Dave A.)
- gluonraid3: 4 partitions 54TB each
- New nodes gluon112-gluon119, gluonraid3
- New firmware installation status
- Busy testing with fADC125
- High Intensity Running
Recharge Wednesday: Ice Cream Novelties
Attendees: David L. (chair), Simon T., Sergey F., Carl T., Naomi J., Dave A., Bryan M., Curtis M., Sean D.
- Graham has a new RAV4 and it is apparently really cool
New equipment installation
- New nodes gluon112-gluon119, gluonraid3 are installed and available for use
- Simon and David confirmed that REHL6u8 binaries are compatible with RHEL6u5 used on other nodes
- Dave A. installed and configured IB switch as well as connections to all nodes.
- He also installed and configured 10Gbit ethernet on gluon113 and gluon115 for testing as SEB
- gluonraid3 also ready for testing as ER
gluonraid3 use for raw data
- David presented a possible new scenario for dataflow through the new gluonraid3 disk on to tape
- System will be changed from what was previously used to prevent reading and writing to the same partition simultaneously
- Each run will be written to a different partition so that data can be moved to silo in a timely matter so that the offline analysis system can get access to it
- System will rely on a set of symbolic links to point to the current, previous, and next partitions to be written to.
- Some coordination between DAQ system and cronjob will be required since both will need to advance links at various points in time
- Carl suggested a file lock be used
- David will work on the scripts incorporating this and other feedback
- Currently still an issue with getting all trigger module firmware versions and driver library versions lined up. This is preventing the whole DAQ system from being used.
- Unable to continue testing fADC125 busy until this is resolved.
- It was agreed that with only ~1 month before the run starts and the new fADC125 firmware not fully tested, that this needs to be considered high priority. It will be brought up at the run planning meeting Thursday morning.
- Newest fADC125 suppresses some event headers
- Unclear if all Event Headers and Trigger Time words are dropped or if the first in the block are retained
- David and Naomi will each ask Cody for clarification
- Will likely require change to offline parsing code. Need data file with new format to fully develop this.
- High intensity running is still being studied
- Current extrapolations point to 4 crates as being the most concerning (Tagger and TOF crates)
- Detector groups notified and work is being done to identify better configurations that will cut out useless hits
- High Intensity testing planned for early in the Fall run to give better extrapolation
- Dave A. has a couple of 10Gbit ethernet cards that can be installed in a ROC.
- We agreed to install one in one of the CDC crates and use the second (unused) 10Gbit uplink in the switch in the hall to test it. This can be instaled next week when the magnet is down.
CAEN1290 Non-linear corrections in firmware
- Sean brought up the situation with the CAEN1290 modules having a non-linear correction to the timing built into the firmware that was tuned for a slightly different clock frequency than what we use.
- Ed J. recently posted a document describing a study he has made of the issue.
- There is a question as to whether we should try loading a new set of corrections based on our clock freq. into the firmware or turn off the non-linear corrections altogether and do them offline as part of the calibration phase.
- Ed is out of town on vacation at the moment, but may be back next week. We decided to ask Beni and Ed to advise us on this since they are most familiar with the issue.