Remove usages of LegacyClusterTaskResultActionListener #83784
Labels
:Distributed Coordination/Cluster Coordination
Cluster formation and cluster state publication, including cluster membership and fault detection.
>refactoring
Team:Distributed (Obsolete)
Meta label for distributed team (obsolete). Replaced by Distributed Indexing/Coordination.
Comments
DaveCTurner
added
>refactoring
:Distributed Coordination/Cluster Coordination
elasticmachine
added
the
Team:Distributed (Obsolete)
|
Pinging @elastic/es-distributed (Team:Distributed) |
DaveCTurner
added a commit
to DaveCTurner/elasticsearch
that referenced
this issue
Feb 10, 2022
DaveCTurner
added a commit
to DaveCTurner/elasticsearch
that referenced
this issue
Feb 10, 2022
DaveCTurner
added a commit
to DaveCTurner/elasticsearch
that referenced
this issue
Feb 10, 2022
Today the `MasterService` permits clients to submit a batch of tasks which it guarantees to execute together, but the only place that this functionality is used in production code is for completing an election. It was done this way so that each join could succeed or fail independently, but since elastic#83562 we can track the status of joins through to completion without needing them all to be separate tasks. This commit introduces a `JoinTask` which represents the whole batch of joins as a single task. It also gives us a place to hang the strange `_FINISH_ELECTION_` task that was used to flag whether a batch was an election-completing batch or not. Relates elastic#83784
This was referenced Feb 10, 2022
DaveCTurner
added a commit
that referenced
this issue
Feb 11, 2022
DaveCTurner
added a commit
that referenced
this issue
Feb 11, 2022
elasticsearchmachine
pushed a commit
that referenced
this issue
Feb 11, 2022
Today the `MasterService` permits clients to submit a batch of tasks which it guarantees to execute together, but the only place that this functionality is used in production code is for completing an election. It was done this way so that each join could succeed or fail independently, but since #83562 we can track the status of joins through to completion without needing them all to be separate tasks. This commit introduces a `JoinTask` which represents the whole batch of joins as a single task. It also gives us a place to hang the strange `_FINISH_ELECTION_` task that was used to flag whether a batch was an election-completing batch or not. Relates #83784
DaveCTurner
added a commit
to DaveCTurner/elasticsearch
that referenced
this issue
Feb 11, 2022
This was referenced Feb 11, 2022
elasticsearchmachine
pushed a commit
that referenced
this issue
Feb 11, 2022
elasticsearchmachine
pushed a commit
that referenced
this issue
Feb 11, 2022
DaveCTurner
added a commit
to DaveCTurner/elasticsearch
that referenced
this issue
Feb 18, 2022
DaveCTurner
added a commit
to DaveCTurner/elasticsearch
that referenced
this issue
Feb 22, 2022
DaveCTurner
added a commit
that referenced
this issue
Mar 2, 2022
DaveCTurner
added a commit
that referenced
this issue
Mar 14, 2022
DaveCTurner
added a commit
to DaveCTurner/elasticsearch
that referenced
this issue
Apr 19, 2022
DaveCTurner
added a commit
that referenced
this issue
Apr 25, 2022
DaveCTurner
added a commit
to DaveCTurner/elasticsearch
that referenced
this issue
Apr 26, 2022
DaveCTurner
added a commit
that referenced
this issue
Apr 26, 2022
DaveCTurner
added a commit
to DaveCTurner/elasticsearch
that referenced
this issue
May 4, 2022
DaveCTurner
added a commit
to DaveCTurner/elasticsearch
that referenced
this issue
May 5, 2022
elasticsearchmachine
pushed a commit
that referenced
this issue
May 5, 2022
DaveCTurner
added a commit
that referenced
this issue
May 5, 2022
|
Hi, @DaveCTurner. It looks like no one is assigned for the last item "to push o.e.cluster.ClusterStateTaskListener#clusterStateProcessed() down to o.e.cluster.ClusterStateUpdateTask". As a first time contributor, I believe I could do it. |
|
Thanks for the offer @PDTCCLF. This item is blocked on other work in the list but when it becomes unblocked it will only take us a few minutes to do so it's not something that will need external help. |
elasticsearchmachine
pushed a commit
that referenced
this issue
Aug 18, 2022
Also removes the now-unused legacy method `ClusterStateTaskListener#onClusterStateProcessed`. Closes #83784
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Description
In #83562 we introduced a new API for better batching of master tasks, and also the
LegacyClusterTaskResultActionListenerto adapt old implementations to the new API as a temporary measure. This issue tracks the work to remove this adapter.o.e.cluster.LocalMasterServiceTaskando.e.cluster.ClusterStateTaskExecutor#unbatched()(Remove LegacyCTRAListener usage in MasterService #83796)o.e.cluster.coordination.NodeRemovalClusterStateTaskExecutor(Remove LegacyCTRAListener usage in NodeRemovalCSTE #83797)o.e.cluster.coordination.JoinTaskExecutor(Submit batches of joins as single tasks #83803)o.e.cluster.metadata.MetadataIndexStateService(Remove LegacyCTRAListener from open-indices batching #83807)o.e.cluster.metadata.MetadataMappingService(Remove LegacyCTRAListener from MetadataMappingService #83811)o.e.cluster.service.MasterServiceTests(Remove LegacyCTRAListener from MasterServiceTests #83840)o.e.cluster.action.shard.ShardStateAction(Remove LegacyCTRAListener from ShardStateAction #83842)o.e.action.admin.indices.create.AutoCreateAction(Remove LegacyCTRAL from AutoCreateAction #84170)o.e.action.admin.indices.rollover.TransportRolloverAction(Remove LegacyCTRAL from TransportRolloverAction #84166)o.e.action.admin.cluster.desirednodes.DesiredNodesClusterStateTaskExecutor(Remove LegacyCTRAL from desired nodes impl #85984)o.e.snapshots.SnapshotsService(Remove LegacyCTRAL from SnapshotsService #86178)o.e.ingest.IngestService(Remove LegacyCTRAL from IngestService #86446)o.e.cluster.metadata.MetadataIndexTemplateService(Remove LegacyCTRAL from MetadataIndexTemplateService #86459)o.e.cluster.metadata.MetadataUpdateSettingsService(Use custom task instead of generic AckedClusterStateUpdateTask #88643)o.e.xpack.ilm.IndexLifecycleRunner(Remove LegacyClusterTaskResultActionListener #89459)o.e.cluster.ClusterStateTaskListener#clusterStateProcessed()down too.e.cluster.ClusterStateUpdateTasksince it only makes sense for unbatched tasks (Remove LegacyClusterTaskResultActionListener #89459)The text was updated successfully, but these errors were encountered: