[2i2c-uk:lis] Enable profileList #3308
Closed
Conversation
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
|
Merging this PR will trigger the following deployment actions. Support and Staging deployments
Production deployments
|
Comment on lines
50
to
62
| - display_name: "Small: ~512 MB RAM / ~0.5 CPU" | ||
| slug: mem_512m | ||
| default: true | ||
| kubespawner_override: | ||
| # increase as requested via https://2i2c.freshdesk.com/a/tickets/1066 | ||
| mem_guarantee: 512M | ||
| mem_limit: 1G | ||
| - display_name: "Large: ~1 GB RAM / ~0.5 CPU" | ||
| slug: mem_1g | ||
| kubespawner_override: | ||
| # increase as requested via https://2i2c.freshdesk.com/a/tickets/1066 | ||
| mem_guarantee: 1G | ||
| mem_limit: 2G |
There was a problem hiding this comment.
I'd like to avoid use of relative words like "small" and "large" because it will make communities use these words that we won't understand without reading up on their configuration first.
I think it would make sense for them to use the existing 4 CPU / 32 GB nodes with node resource allocation script generated requests/limits where the limit is 2x the request. Is the script able to do that atm? Hmmm...
There was a problem hiding this comment.
I actually did a deploy with just the sizes as names and it felt very mathematical 😅
I think it would make sense for them to use the existing 4 CPU / 32 GB nodes with node resource allocation script generated requests/limits where the limit is 2x the request. Is the script able to do that atm? Hmmm..
I am not familiar with what the script can do yet 😬
There was a problem hiding this comment.
Neither am I and I failed to quickly whip up a suggested alternative - so lets go with anything for now and iterate!
The script currently only provides requests and limits being set to the same value, so nothing matches exactly what we want with limits higher than requests.
consideRatio
approved these changes
Oct 24, 2023
|
@consideRatio, thank you! I was thinking just now, if we are moving to profileLists, why not add their initial memory guarantee as another, even smaller option (256MB)? |
|
@GeorgianaElena I suggest we go for #3308 (comment) and ask them for feedback |
Co-authored-by: Erik Sundell <[email protected]>
consideRatio
approved these changes
Oct 24, 2023
UpdateWaiting for feedback before merging. The profile list looks like: 
|
|
I will close this PR to de-clutter and will open up again if/when agreement about this one. |
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
Update: waiting for feedback in https://2i2c.freshdesk.com/a/tickets/1066 before merging.
Follow-up to #3302, also for https://2i2c.freshdesk.com/a/tickets/1066.
Matthew from LIS requested a guarantee of 1GB as they are still experiencing crashes, even after the increase to 512GB of guarantee from #3302.
This PR enables profileLists for the lis hub, adding the new guarantee and reducing the limit of the initial one of 512G one from 2GB to 1GB.
I think it would be useful for them to enable https://github.com/jupyter-server/jupyter-resource-usage so they can better understand they mem needs, especially since from the grafana below, it looks like only a few users actually required around or more than 1GB.
This is also why the new server size option, instead of increasing it again.