AP Pipeline Meeting, 2023-07-17

2pm Pacific Time

Zoom: https://washington.zoom.us/j/98846655020

attending:

• Eric Bellm
• Krzysztof Findeisen 
• Erin Howard 
• Kenneth Herner 
• Meredith Rawls 

regrets

• Ian Sullivan (vacation)

Topics for discussion:

• Project updates (Eric Bellm , Ian Sullivan ):
  • nothing notable
• Auxtel processing status (all):
  • run coming up this week–going try running diffim, again
• Pipelines Infrastructure (Krzysztof Findeisen , John Parejko ):
  • DM-40046 - Getting issue details... STATUS – being fixed this afternoon with Erin
• Precursor processing & Campaign Management ( Kenneth Herner , Erin Howard):
  • Ken: following a thread about job overhead in #ops-usdf in condor, panda, interactive SLURM.
    • otherwise trying to complete his ap_verify run comparison script.
  • Erin
    • APDB filled up again. .
    • has been cleaning the up the APDB–we're not sure why the APDB is filling when the APDB is not being used
    • it filled up when Erin had not made any large runs
    • Krzysztof asked Dan Speck on PP channel if there was anything else using the Postgres db or the partition, so we don't know what's going on
      • Dan will look into it further when he gets back from vacation
    • HSC runs are ~40 GB, HiTS ~3 GB. 
    • some open questions about variance from run to run, and why the db is at capacity when the schemas seem smaller
    • per previous discussion, keeping ~2 older monthly runs seems reasonable–Erin is rotating/deleting as requested
    • 1 TB hopefully gives us enough overhead?
  • Erin
    • otherwise, started Saha Bulge run
• Image differencing algorithms, DCR  ( Bruno Sanchez , Ian Sullivan ):
  • out today
• alert distribution (Brianna Smart , Eric Bellm ):
  • out today
• transinet (Nima Sedaghat Harshit Rai , Eric Bellm ):
  • out today
• Solar-system processing (Ari Heinze ):
  • out today
• Review CI (https://usdf-rsp-dev.slac.stanford.edu/chronograf/):
  • a one un-associated DIASource change in HiTS, which we hypothesize is DM-39733 - Getting issue details... STATUS
  • but actually it's too early, maybe DM-39804 - Getting issue details... STATUS ?  but we don't see any speedups in the timing, so maybe it's not being run yet?
  • in #dm-jenkins it looks like 7.0 was available on the 13th, and https://lsstc.slack.com/archives/C2JPMCF5X/p1689621951635779 discusses a number of DRP regressions at the same time.
• Review outstanding action items
• QA meeting September:
  • Eric on detection thresholds/fakes forced photometry?
  • streak masking/long trail rejection (Bri & Meredith)
• AOB
  • Eric will be on vacation next two weeks, Ian will run the meeting

Action Items