2.13.6

Changeset in respect to 2.13.5:

• handle missing per FED deadtime monitoring in Backkpressure-from-EvB analysis (GitHub issue 236)
• fix of backpressure reporting of FED-deadtime-due-to-DAQ problem
• fix of backpressure and problematic-FED reporting in Backpressure-from-EvB problem
• fix of affected-partition reporting for Corrupted-data problem
• handle lack of instantaneous deadtime in monitoring

2.13.4

• implemented #233: holdoff (duration of condition and start of run effects) for HLT bandwidth modules

2.13.3

• fixed context key in OutOfSequenceData module
• removed additional note in HltCpuLoad module

2.11.0 - causality relations

Key improvements:

• better selection of root problem (what shifter sees as a suggestion in the end)
• updated API for Run Logger - more concise, brief descriptions of a problem instead of long, full descriptions
• updated API for Run Logger - only one root problem supplied
• final suggestion for operator is now also available in the browser
• better reporting of many feds - if possible ranges will be used eg. 100-105 instead of 100, 101, 102, ...

2.10.16

• fixed #226: machine development / technical stop (LHC machine modes) are now treated the same way as interfill beam modes far before stable beams (no maximum fraction of FUs in cloud mode is imposed)

2.10.15 Filter upgraded feds

In temporary LM that was introduced due to lacking per fed deadtime information we are now filtering upgraded feds. This is because this check doesn't apply for legacy FEDs.

2.10.14 - per-subsystem threshold conf. for fixing-soft-error

Thresholds might be now individually configured. See details in #223

2.10.13 - improve rate out of range

LM Rate out of range will now check the avg RU rate only during stable beams and when NoRate LM is not satisfied. Description will now include the rate:

L1 rate {{RATE}} is out of expected range (" + min / 1000f + " - " + max / 1000f + " kHz)

2.10.12 - fix counting of subsystems going to fixing-soft-error

Bugfix to resolve counting of subsystems going to fixing-soft-error. This resulted in descriptions like

Level zero repeatably in FixingSoftError due to subsystem(s) [TRACKER 9 time(s), TRACKER 10 time(s), TRACKER 11 time(s), TRACKER 12 time(s)]

ECAL and ES special instructions for LHC clock mode unstable

2.10.4

Includes ECAL and ES special instructions for LHC clock mode unstable