|
Size: 441
Comment:
|
← Revision 10 as of 2018-09-21 10:59:55 ⇥
Size: 1536
Comment:
|
| Deletions are marked like this. | Additions are marked like this. |
| Line 11: | Line 11: |
| Osamu, Takahiro, Keiko, Oshinosan, Keith, Rolf | |
| Line 15: | Line 15: |
| Line 17: | Line 16: |
| * Report on IRIG-B issue | * Report on IRIG-B issue and CPU max * presentation file is [[https://gwdoc.icrr.u-tokyo.ac.jp/cgi-bin/DocDB/ShowDocument?docid=9017|here]]. |
| Line 19: | Line 19: |
| * A file is [[https://gwdoc.icrr.u-tokyo.ac.jp/cgi-bin/DocDB/ShowDocument?docid=9017|here]]. | * [[https://gwdoc.icrr.u-tokyo.ac.jp/cgi-bin/DocDB/ShowDocument?docid=9018|viewgraph]]. * KAGRA's DMT and DQ Flags * LIGO's svn? |
| Line 23: | Line 25: |
== CPU max == * epics channels to store => DAQ flow * general epics channel => normal TCP/IP flow * RCG 3.1 improves the double traffic * Also slower than 3GHz CPU is not good * slot of PCIE card - one stop card matters * when we put a interface card of the raiser card to closer to mother board card was not good - LIGO: it shouldn't matter * Priority of riser cards may be necessary (not much related to the current cpu issue) * epics gateway (machine) may be necessary to collect the epics information and reduce the redundant traffic == broadcaster == * dedicated network from bcst low latency * not same swich from the control room client machiens' * 1Gb switch is fine * not all chans are sent to dmt == svn == * svn web interface changed * freeze the svn at RCG 3.5 for O3 and move to git * svn version is updated - Keith will send the info |
LIGO-KAGRA CDS meeting - 6/21 8am (JST), 6/20 6pm (LLO)
Connection Info
Teamspeak (password protected page. KAGRA wiki account is necessary to view.)
Participants
Osamu, Takahiro, Keiko, Oshinosan, Keith, Rolf
Agenda
- Report on IRIG-B issue and CPU max
presentation file is here.
- Report on k1bcst0
- KAGRA's DMT and DQ Flags
- LIGO's svn?
Minutes
CPU max
epics channels to store => DAQ flow
general epics channel => normal TCP/IP flow
- RCG 3.1 improves the double traffic
- Also slower than 3GHz CPU is not good
- slot of PCIE card - one stop card matters
- when we put a interface card of the raiser card to closer to mother board card was not good - LIGO: it shouldn't matter
- Priority of riser cards may be necessary (not much related to the current cpu issue)
- epics gateway (machine) may be necessary to collect the epics information and reduce the redundant traffic
broadcaster
- dedicated network from bcst low latency
- not same swich from the control room client machiens'
- 1Gb switch is fine
- not all chans are sent to dmt
svn
- svn web interface changed
- freeze the svn at RCG 3.5 for O3 and move to git
- svn version is updated - Keith will send the info