Avoid Needless Cache Status Fetches in SearchableSnapshotAllocator #66433
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
We shouldn't fetch cache status if no allocation is possible to begin with. Also, this surfaced an issue with using the `Client` to `reroute` since that won't retry stale shards (failed the invalid license IT for example) so I moved to using the `RerouteService` like we do in the `GatewayAllocator`. (Plus, dried up one method that was 100% the same as in the replica allocator)
original-brownbear
added
>non-issue
:Distributed Coordination/Snapshot/Restore
elasticmachine
added
Team:Distributed (Obsolete)
|
Pinging @elastic/es-distributed (Team:Distributed) |
1 similar comment
|
Pinging @elastic/es-distributed (Team:Distributed) |
| @@ -44,7 +44,7 @@ | |||
| explanations = RoutingExplanations.readFrom(in); | |||
| } | |||
|
|
|||
| public ClusterRerouteResponse(boolean acknowledged, ClusterState state, RoutingExplanations explanations) { | |||
| ClusterRerouteResponse(boolean acknowledged, ClusterState state, RoutingExplanations explanations) { | |||
There was a problem hiding this comment.
Just a revert of making this public yesterday now that it's not needed for tests any longer.
henningandersen
approved these changes
Dec 16, 2020
| // pre-check if it can be allocated to any node that currently exists, so we won't list the cache sizes for it for nothing | ||
| // TODO: in the following logic, we do not account for existing cache size when handling disk space checks, should and can we | ||
| // reliably do this in a world of concurrent cache evictions or are we ok with the cache size just being a best effort hint | ||
| // here? | ||
| Tuple<Decision, Map<String, NodeAllocationResult>> result = canBeAllocatedToAtLeastOneNode(shardRouting, allocation); | ||
| Tuple<Decision, Map<String, NodeAllocationResult>> result = ReplicaShardAllocator.canBeAllocatedToAtLeastOneNode( |
There was a problem hiding this comment.
Can we add a test that we do not trigger any reads or reroutes when deciders say no?
There was a problem hiding this comment.
Sure thing, I pushed 0cb0017 :)
|
Thanks Henning! |
original-brownbear
added a commit
to original-brownbear/elasticsearch
that referenced
this pull request
Dec 16, 2020
…lastic#66433) We shouldn't fetch cache status if no allocation is possible to begin with. Also, this surfaced an issue with using the `Client` to `reroute` since that won't retry stale shards (failed the invalid license IT for example) so I moved to using the `RerouteService` like we do in the `GatewayAllocator`. (Plus, dried up one method that was 100% the same as in the replica allocator)
original-brownbear
added a commit
that referenced
this pull request
Dec 16, 2020
…66433) (#66444) We shouldn't fetch cache status if no allocation is possible to begin with. Also, this surfaced an issue with using the `Client` to `reroute` since that won't retry stale shards (failed the invalid license IT for example) so I moved to using the `RerouteService` like we do in the `GatewayAllocator`. (Plus, dried up one method that was 100% the same as in the replica allocator)
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.
We shouldn't fetch cache status if no allocation is possible to begin with.
Also, this surfaced an issue with using the
Clienttoreroutesince thatwon't retry stale shards (failed the invalid license IT for example) so I moved
to using the
RerouteServicelike we do in theGatewayAllocator.(Plus, dried up one method that was 100% the same as in the replica allocator)