DDO-3917-scale-down-buffer-pools #374
Conversation
| resourceConfigName: "cwb_ws_resource_fiab_v8" | ||
| - poolId: "cwb_ws_quality_v8" | ||
| size: 5000 | ||
| resourceConfigName: "cwb_ws_resource_quality_v8" |
There was a problem hiding this comment.
I think each resourceConfigName corresponds to a file, should those be deleted too? https://github.com/DataBiosphere/terra-resource-buffer/blob/master/src/main/resources/config/tools/resource-config/cwb_ws_resource_quality_v8.yml
There was a problem hiding this comment.
I went ahead and removed all the config files, checking around to make sure this is a safe-ish move though.
|
|
hey folks! can I get some thumbs to push this one through? |
Unfortunately, Buffer never got ported to Bees (we would have had to solve of workspace pooling across multiple buffers as well as buffer orphaned workspaces). I haven't contributed to this repo before, but I expect we can just roll it back |
|
just realized I might have answered a different question than you were asking, (do the new pools work vs will buffer crash) @jgainerdewar mentioned she just tested the new pools in a bee |
Looking at the last few weeks of Buffer activity, it looks like we're quite over provisioned on projects, while technically we're not paying for these, it looks like the number of projects is breaking the billing apis via timeout and thus our ability to track costs across the org.
This PR removes what looks like unused pools as well as scales down the big used pools to ~50% of their peak usage.
https://grafana.dsp-devops.broadinstitute.org/d/dcvlwY2nz/resource-buffer-service?orgId=1&var-cluster=terra-integration&var-namespace=terra-tools&var-project=&var-container=All&viewPanel=4&refresh=10s&from=now-30d&to=now