Outage tasks for 2023-09-14

[kltm]

• Send reminder email to go-consortium mailing list
• Run noctua-models-migration pipeline for (@vanaukenk) (automated to day before) to get report
  http://skyhook.berkeleybop.org/noctua-models-migrations/reports/
• Refresh minerva code with latest from minerva master
• Update NEO
• Load YeastPathways models as development @dustine32 YeastPathways models to Noctua prod for geneontology/pathways2GO#275 noctua-models#257
• Run replaced_by term update on blazegraph (@vanaukenk)
• Cycle minerva to get latest ontology
• Prep ticket for next outage

[kltm]

@vanaukenk migration mock run nom complete.

[vanaukenk]

@kltm
I've double-checked the diff and all looks consistent with recent ontology changes.
The only models affected are MGI gene-centric models.
Thx.

[suzialeksander]

@vanaukenk are you aware we are trying to sneak the YeastPathways models in with state = development? Just in case you see a ton of new models or something, note that this is expected- or please correct me if I have my timings off @dustine32

[dustine32]

Correct @suzialeksander and thank you for mentioning this here! I'm currently trying to see about fixing geneontology/pathways2GO#276 before the load to Prod (it looks like just commenting out a single line but I wanna test some models locally first). If I'm unable to get a fixed set of models by tomorrow, should we hold off on the prod load until the next maintenance window? Or should we just push what we have?

[suzialeksander]

@dustine32 as long as they're state=dev, I'm happy to push it anyway. Curation-wise, removing that edge is fairly minor to the other changes needed.

[kltm]

Just to clarify my understanding from poking around a little and the vibe:

• this is for a handful of models (sjcarbon@moiraine:~/local/src/git/noctua-models/models[yeast-pathways-20230811]$:) ls | grep -i "yeastpathways" | wc -l ->220)
• any further changes to models are minor
• any further changes will be done manually in the interface by @suzialeksander, w/o a reload later on
• all models are dev and will accordingly not appear in public interfaces

Does this sound right?

[vanaukenk]

Thank you @suzialeksander @dustine32 @kltm

I wasn't aware that the yeast pathways models were going onto Noctua production for this outage. Keep us posted if anything changes.

[dustine32]

@kltm Yeah, this only affects 220 models and they'll go in as state=development.

[vanaukenk]

@suzialeksander

Are you planning to do some QC checks on the yeast pathway models after tonight's outage? If so, do you want me to hold off on sending out the all clear email until you are finished?

[suzialeksander]

@vanaukenk I was planning on checking some of these, but I don't think there's any reason to hold off on the general email if it's just these models that have any issues. We plan on curating each one manually so tonight's checks will not be particularly in-depth.