Details
-
Type:
Story
-
Status: Done
-
Resolution: Done
-
Fix Version/s: None
-
Component/s: None
-
Labels:None
-
Story Points:1.4
-
Epic Link:
-
Team:SQuaRE
-
Urgent?:No
Description
Today I dig into why the stellar locus metric wPerp hasn't been appearing on our RC2 dashboard lately, despite Brock Brendal [X]'s reports that for at least 9615 and 9697 the metric measurement succeeded.
Turns out the it's been appearing as tellar_locus_width_wPerp.
The last metric name with an s was Jan 13. the first metric appearing as tellar was Jan 20.
Only squash-demo with the monthlies seems to be affected. squash-prod with the nightlies seems fine. That is pretty confusing to me.
Attachments
Issue Links
- relates to
-
- Won't Fix
Activity
Thanks for reporting.
That's my fault, when I put together the notebook to sync data between SQuaSH and InfluxDB I didn't include this fix. It only affects squash-demo which requires this manual step.
I'll fix the notebook and try to manually fix the squash-demo database as well.
I confirm that the version of the notebook I'm running doesn't have the fix described in DM-31566. Note the wrong metric name tellar_locus_width_wPerp.
Sending InfluxDB line for job 81077 dataset HSC_RC2.
|
pipe_analysis,tract=9615,env_name=ldf,dataset=HSC_RC2,skymap=hsc_rings_v1,filter=u,dataset_repo_url=https://community.lsst.org/t/shared-gen3-data-repositories-ready-for-some-use/4845,butler_generation=Gen3 tellar_locus_width_wPerp=11.1745,run_id="DM-33402",run_url="[DM-33402](https://jira.lsstcorp.org/browse/DM-32402)",version_tag="w_2022_04",squash_id=81077,squash_url="[81077](https://squash-restful-api.lsst.codes/job/81077)",timestamp=1645068944000000000 1642682340000000000
|
I'm about to begin RC2 w16, so we'll see if the wPerp jobs start appearing as they should!
This can be closed now.
I also removed the following query from the Photometry: stellar_locus_width_wPerp (stellar locus width) panel which is not required anymore.
SELECT "tellar_locus_width_wPerp" as "tellar_wPerp" FROM "squash-demo"."autogen"."pipe_analysis" WHERE time > :time_start: AND time < :upperDashboardTime: AND "dataset"=:dataset: AND "tract"=:tract: AND "butler_generation"=:butler_gen: ;
|
I reported the same issue back in August. I think either Angelo or Simon fixed it, but I can't find the message where they reported the solution (I believe it was a bad regex).