set timestamp in reconstruction#8341
Conversation
|
FT0 should be reconstructed in pass4 with this PR to have it calibrated |
shahor02
left a comment
shahor02
left a comment
There was a problem hiding this comment.
Merging, but you should consider switching to DPL CCDB support.
| CalibObjWithInfoType result; | ||
| auto clName = o2::utils::MemFileHelper::getClassName(calibrationObject); | ||
| auto flName = o2::ccdb::CcdbApi::generateFileName(clName); | ||
| stopping = 999999999999; |
There was a problem hiding this comment.
Actually, before I merge: what is the reason to set inf. end validity for every object? Overlapping objects create overhead in the query. Cannot you define one default object with large validity and upload calibration results with reasonable validity, e.g. end timestamp of the calibration data + some reasonable margin (e.g. 1 day) ?
There was a problem hiding this comment.
The reason of so large validity is that if to use offsets from previous runs if something went wrong.
There was a problem hiding this comment.
But how old calibration you are ready to accept? I can understand a few days old, otherwise, why don't you put some reasonable default only once?
With the current infinite settings, all your objects will overlap.
There was a problem hiding this comment.
OK, I will set validity as last timestamp + 1 day and upload default calibration with infinite settings - as you suggest
AllaMaevskaya
force-pushed
the
setOffsets
branch
from
2fe5d30 to
72ec61c
Compare
shahor02
left a comment
shahor02
left a comment
There was a problem hiding this comment.
Thanks! Merging since the tests were passed before trivial change.
3 participants
No description provided.