fix(scheduler): Add grace period for model unload #5965
Merged
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Outdated suggestions cannot be applied.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:
Currently when a model is scaling down (reducing the number of replicas), and while we make sure that we issue a remove route to the server containing the replica before unloading the replica, we found issues where there are still race conditions between envoy actually applying the cluster change and the model replica unloading.
This PR introduces a grace period that is applied before any unload to a particular replica to guard against this race condition. Typically envoy cluster updates take ~500ms and we set the default grace period to 2s.
The value can be changed via envar on seldon-agent
SELDON_UNLOAD_GRACE_PERIOD_SECONDS
.