June 27, 2018 online meeting minutes: Difference between revisions
Jump to navigation
Jump to search
No edit summary |
No edit summary |
||
| (8 intermediate revisions by the same user not shown) | |||
| Line 1: | Line 1: | ||
present: Sergey Boyarinov, | present: Sergey Boyarinov, Stepan, Ben, Hai, Ed, Graham | ||
We discussed a progress in DAQ/Trigger upgrades for upcoming run. | We discussed a progress in DAQ/Trigger upgrades for upcoming run. | ||
FADC bit packing: on last meeting we planed that in the end of June we may have something to test, and will try to get one FADC250 version for everybody | * FADC bit packing: on last meeting we planed that in the end of June we may have something to test, and will try to get one FADC250 version for everybody; Hai finished simulation and passed to Ed; Ed works with Ben and expect first version for testing by July 3; FADC busy may not be there in first version, will be implemented soon after; Bryan Moffit has to change library to handle a couple of extra control bits; Sergey has to change rol2 to understand packed data; have to tell Gagik to put decoding in | ||
DC road finder: | * DC road finder: last time we discussed to have first version mid of July; Sergey will generate dict today - just 6 columns | ||
TI/TD/TS firmwares and libraries: Hall D is finished testing, observed significant improvements in stability, some issues may still exist in fiber length measurements; | * TI/TD/TS firmwares and libraries: Hall D is finished testing, observed significant improvements in stability, some issues may still exist in fiber length measurements; will make standalone measurement procedure, use fixed values in download; will put 8.1 into FC first and test with Cole and performance test | ||
CLAS12 Event Recorder testing with different raid configurations, number of partitions and number of streams on clondaq6 with Paul Letta; found that raid0 single partition has highest performance, raid5 single partition works as good as raid0s; multi-partition does not give any advantages and requires more complex logic; 4 streams seems optimal; inserting data into ET grand station data rate about 2.5GB/sec was achieved on empty disk and free clondaq6 [https://clonwiki0.jlab.org/wiki/clondocs/Docs/raid5_4stream_1partition_nextday.png | * Procurements: TD boards, VTP repairs 1 approved, 3 more to go; VXS crates, fileserver+3 IOCs, CPUs, [DSCs], SDs, oldTD->TI to go; make sure QSFP transceivers are good | ||
* Online farm: Graham reviewed Sergey's proposal, Sergey have to add drawing, justify 40GBit switch, can it be part of existing farm; Sergey and Graham will discuss it afternoon | |||
* CAEN TDC calibration: Sergey will do it during summer | |||
* CLAS12 Event Recorder testing with different raid configurations, number of partitions and number of streams on clondaq6 with Paul Letta; found that raid0 single partition has highest performance, raid5 single partition works as good as raid0s; multi-partition does not give any advantages and requires more complex logic; 4 streams seems optimal; inserting data into ET grand station data rate about 2.5GB/sec was achieved on empty disk and free clondaq6 [https://clonwiki0.jlab.org/wiki/clondocs/Docs/raid5_4stream_1partition_nextday.png (runcontrol screenshot)]. | |||
* finish checkin in GIT CLAS12 CODA software (CLON_PARMS have to be cleaned up) | |||
Latest revision as of 10:03, 27 June 2018
present: Sergey Boyarinov, Stepan, Ben, Hai, Ed, Graham
We discussed a progress in DAQ/Trigger upgrades for upcoming run.
- FADC bit packing: on last meeting we planed that in the end of June we may have something to test, and will try to get one FADC250 version for everybody; Hai finished simulation and passed to Ed; Ed works with Ben and expect first version for testing by July 3; FADC busy may not be there in first version, will be implemented soon after; Bryan Moffit has to change library to handle a couple of extra control bits; Sergey has to change rol2 to understand packed data; have to tell Gagik to put decoding in
- DC road finder: last time we discussed to have first version mid of July; Sergey will generate dict today - just 6 columns
- TI/TD/TS firmwares and libraries: Hall D is finished testing, observed significant improvements in stability, some issues may still exist in fiber length measurements; will make standalone measurement procedure, use fixed values in download; will put 8.1 into FC first and test with Cole and performance test
- Procurements: TD boards, VTP repairs 1 approved, 3 more to go; VXS crates, fileserver+3 IOCs, CPUs, [DSCs], SDs, oldTD->TI to go; make sure QSFP transceivers are good
- Online farm: Graham reviewed Sergey's proposal, Sergey have to add drawing, justify 40GBit switch, can it be part of existing farm; Sergey and Graham will discuss it afternoon
- CAEN TDC calibration: Sergey will do it during summer
- CLAS12 Event Recorder testing with different raid configurations, number of partitions and number of streams on clondaq6 with Paul Letta; found that raid0 single partition has highest performance, raid5 single partition works as good as raid0s; multi-partition does not give any advantages and requires more complex logic; 4 streams seems optimal; inserting data into ET grand station data rate about 2.5GB/sec was achieved on empty disk and free clondaq6 (runcontrol screenshot).
- finish checkin in GIT CLAS12 CODA software (CLON_PARMS have to be cleaned up)