Skip to content

[REMI] Move pulse length warning from kernel to after separation

Philipp Schmidt requested to merge fix/REMI-pulse-len-warning into master

Description

There was an issue yesterday night with the REMI reconstruction somewhat silently "crashing" due to excessive warning prints over different pulse lengths. The crashing part actually originates from an outdated version of jupyter_client (fixed here for pycalibration proper) due to this notebook using is own environment with proprietary libraries. Still, a per-train warning in the kernel seems excessive as this case will most likely occur for all trains at the same time due to a machine misconfiguration. Hence I removed this warning and instead do some post-separation statistics. Interestingly this also split the problem into two problems:

  • Different pulse lengths in between trains, which seems less likely to be a problem, e.g. through alternating fill patterns
  • Different pulse lengths in a single train (or more) such as the case triggering this issue. This seems much less likely to be intended, as it can easily mess up the per-pulse signals.

How Has This Been Tested?

Run without any issues: p2979, r10 Run triggering this: p2979, r18

Types of changes

  • Bug fix (non-breaking change which fixes an issue)

Reviewers

@ahmedk (Sorry for the call!)

Merge request reports

Legal Notice | Contacts