Muon Week 2017 - Run2 Computing

News

Laurent Aphecetche

19 May 2017

ALIROOT-6564

LV

Bring the treatment at the same level as the one from HV
if there is a LV trip during a run, kill the affected channels for the whole run
code is there (since october last year)
not activated yet (matter of uploading a new recoparam OCDB object)
should we do it now ? yes : no objections ?

HV : the 1400 syndrom

Current HV algorithm clean-up the giant fluctuation around 1400
is then left with 1 perfectly fine HV value
channel declared good for offline
which is obviously wrong

need to change the algorithm
without affecting the rest of the channels
work started but not completed yet

current idea is to compute the fraction of time a channel is below working voltage
then set a threshold on that fraction

Config

Config = set of active buspatches
Buspatches can now be removed "on-the-fly" during a run by the MCH OnCall (See Vincent's talk)
Like HV or LV, a buspatch removed for any length of time during a run will be killed for the full run
i.e. we stay with (no more than) one config per run

change of config is currently not transmitted to OCDB for each run
but only at next pedestal runs
will be fixed

Not done yet and probably never

flag(s) to know, at ESD level, if an event reconstruction was aborted
- we lived w/o them so far, probably can til end of Run2 as well...
remove unused OCDB objects
- would be clean but not really needed actually

Priority is now Run3

only critical bugs will be fixed for Run2
if workaround exist, use that instead of invest time in developping a proper fix
- example of DQM MCH agent to be restarted when config changes a lot

Communication

Private emails are inefficient (for both the sender and the receiver)
- sender can not get a timely answer if receiver not available
- receiver might have to answer several time the same questions

Please use mailing list(s) and/or JIRA for bug reports and questions

e.g. alice-dimuon-offline@cern.ch or alice-dimuon-pwg3@cern.ch

[Your question here]

1

Muon Week 2017 - Run2 Computing News Laurent Aphecetche 19 May 2017