fix(scheduler): Add >0 replica check for server filter scheduling #4928
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
ukclivecox
changed the title
sakoush
reviewed
Jul 10, 2023
| } | ||
|
|
||
| func (r ServerReplicaFilter) Filter(model *store.ModelVersion, server *store.ServerSnapshot) bool { | ||
| return len(server.Replicas) > 0 |
There was a problem hiding this comment.
Could we make this a bit more general to filter out any server with number of replicas < number of model replicas?
Although I am not sure if it can cause any issues with scaling models.
There was a problem hiding this comment.
I think we catch this later though anyway
sakoush
approved these changes
Jul 10, 2023
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Outdated suggestions cannot be applied.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.
What this PR does / why we need it:
At present if a Server resource is created the operator informs the scheduler which creates an embryo Server internal data structure. This is not filled out until the Server actually connects to the scheduler when it comes up. Until then scheduling will usually fail due to an unclear message on sharing=false as this is the default value for the Server internal data structure. This is unintuitive so this PR adds a new filter that checks any server has >0 replicas. If a server is still being created (maybe a big delay due to large image download) this error will be returned if a model can not be scheduled to it.