Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Update index mappings when ccr restore complete #36879

Merged
merged 6 commits into from
Dec 20, 2018

Conversation

Tim-Brooks
Copy link
Contributor

This is related to #35975. When the shard restore process is complete,
the index mappings need to be updated to ensure that the data in the
files restores is compatible with the follower mappings. This commit
implements a mapping update as the final step in a shard restore.

@Tim-Brooks Tim-Brooks added >non-issue v7.0.0 :Distributed Indexing/CCR Issues around the Cross Cluster State Replication features v6.7.0 labels Dec 20, 2018
@Tim-Brooks Tim-Brooks requested a review from ywelsch December 20, 2018 04:27
@elasticmachine
Copy link
Collaborator

Pinging @elastic/es-distributed

@Tim-Brooks
Copy link
Contributor Author

This is an attempt to continue to split the full file restore into smaller pieces. Especially, as we continue to work out how to handle the remote requests to specific nodes.

Copy link
Contributor

@ywelsch ywelsch left a comment

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

@martijnvg can you have a look as well please?

}

@Override
public IndexShardSnapshotStatus getShardSnapshotStatus(SnapshotId snapshotId, Version version, IndexId indexId, ShardId leaderShardId) {
throw new UnsupportedOperationException("Unsupported for repository of type: " + TYPE);
}

private void maybeUpdateMappings(Client localClient, Client remoteClient, Index leaderIndex, IndexSettings followerIndexSettings) {
Copy link
Contributor

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

@martijnvg would it make sense to share this code with the one in ShardFollowTasksExecutor?

Copy link
Member

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Conceptually this code is doing the same as the update mapping code in shard follow task, but the update mapping code is tightly coupled with ShardFollowTasksExecutor and ShardFollowNodeTask. Also the mapping update code is asynchronous over there and here it is synchronous.

What I think is possible, is that factory methods for both ClusterStateRequest and PutMappingRequest are introduced that is then used here and in ShardFollowTasksExecutor.

@Tim-Brooks
Copy link
Contributor Author

@martijnvg @ywelsch I made some changes to share the request creation.

@Tim-Brooks
Copy link
Contributor Author

@ywelsch Made your changes.

Copy link
Contributor

@ywelsch ywelsch left a comment

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

LGTM

@Tim-Brooks Tim-Brooks merged commit 4cd5705 into elastic:master Dec 20, 2018
Tim-Brooks added a commit to Tim-Brooks/elasticsearch that referenced this pull request Dec 20, 2018
This is related to elastic#35975. When the shard restore process is complete,
the index mappings need to be updated to ensure that the data in the
files restores is compatible with the follower mappings. This commit
implements a mapping update as the final step in a shard restore.
Tim-Brooks added a commit that referenced this pull request Dec 21, 2018
This is related to #35975. When the shard restore process is complete,
the index mappings need to be updated to ensure that the data in the
files restores is compatible with the follower mappings. This commit
implements a mapping update as the final step in a shard restore.
import org.elasticsearch.cluster.metadata.MappingMetaData;
import org.elasticsearch.common.xcontent.XContentType;

public final class CcrRequests {
Copy link
Member

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

👍

jasontedor added a commit to ywelsch/elasticsearch that referenced this pull request Dec 21, 2018
* elastic/master: (539 commits)
  SQL: documentation improvements and updates (elastic#36918)
  [DOCS] Merges list of discovery and cluster formation settings (elastic#36909)
  Only compress responses if request was compressed (elastic#36867)
  Remove duplicate paragraph (elastic#36942)
  Fix URI to cluster stats endpoint on specific nodes (elastic#36784)
  Fix typo in unitTest task (elastic#36930)
  RecoveryMonitor#lastSeenAccessTime should be volatile (elastic#36781)
  [CCR] Add `ccr.auto_follow_coordinator.wait_for_timeout` setting (elastic#36714)
  Scripting: Remove deprecated params.ctx (elastic#36848)
  Refactor the REST actions to clarify what endpoints are deprecated. (elastic#36869)
  Add JDK 12 to CI rotation (elastic#36915)
  Improve error message for 6.x style realm settings (elastic#36876)
  Send clear session as routable remote request (elastic#36805)
  [DOCS] Remove redundant ILM attributes (elastic#36808)
  SQL: Fix bug regarding histograms usage in scripting (elastic#36866)
  Update index mappings when ccr restore complete (elastic#36879)
  Docs: Bump version to alpha2 after release
  Enable IPv6 URIs in reindex from remote (elastic#36874)
  Watcher: Remove unused local variable in doExecute (elastic#36655)
  [DOCS] Synchs titles of X-Pack APIs
  ...
jasontedor added a commit to jasontedor/elasticsearch that referenced this pull request Dec 21, 2018
* master: (31 commits)
  Move ingest-geoip default databases out of config (elastic#36949)
  [ILM][DOCS] add extra scenario to policy update docs (elastic#36871)
  [Painless] Add String Casting Tests (elastic#36945)
  SQL: documentation improvements and updates (elastic#36918)
  [DOCS] Merges list of discovery and cluster formation settings (elastic#36909)
  Only compress responses if request was compressed (elastic#36867)
  Remove duplicate paragraph (elastic#36942)
  Fix URI to cluster stats endpoint on specific nodes (elastic#36784)
  Fix typo in unitTest task (elastic#36930)
  RecoveryMonitor#lastSeenAccessTime should be volatile (elastic#36781)
  [CCR] Add `ccr.auto_follow_coordinator.wait_for_timeout` setting (elastic#36714)
  Scripting: Remove deprecated params.ctx (elastic#36848)
  Refactor the REST actions to clarify what endpoints are deprecated. (elastic#36869)
  Add JDK 12 to CI rotation (elastic#36915)
  Improve error message for 6.x style realm settings (elastic#36876)
  Send clear session as routable remote request (elastic#36805)
  [DOCS] Remove redundant ILM attributes (elastic#36808)
  SQL: Fix bug regarding histograms usage in scripting (elastic#36866)
  Update index mappings when ccr restore complete (elastic#36879)
  Docs: Bump version to alpha2 after release
  ...
dnhatn added a commit that referenced this pull request Feb 4, 2019
There are two issues regarding the way that we sync mapping from leader
to follower when a ccr restore is completed:

1.  The returned mapping from a cluster service might not be up to date
as the mapping of the restored index commit.

2. We should not compare the mapping version of the follower and the
leader. They are not related to one another.

Moreover, I think we should only ensure that once the restore is done,
the mapping on the follower should be at least the mapping of the copied
index commit. We don't have to sync the mapping which is updated after
we have opened a session.

Relates #36879
Closes #37887
Tim-Brooks pushed a commit to Tim-Brooks/elasticsearch that referenced this pull request Feb 5, 2019
There are two issues regarding the way that we sync mapping from leader
to follower when a ccr restore is completed:

1.  The returned mapping from a cluster service might not be up to date
as the mapping of the restored index commit.

2. We should not compare the mapping version of the follower and the
leader. They are not related to one another.

Moreover, I think we should only ensure that once the restore is done,
the mapping on the follower should be at least the mapping of the copied
index commit. We don't have to sync the mapping which is updated after
we have opened a session.

Relates elastic#36879
Closes elastic#37887
Tim-Brooks added a commit that referenced this pull request Feb 6, 2019
There are two issues regarding the way that we sync mapping from leader
to follower when a ccr restore is completed:

1.  The returned mapping from a cluster service might not be up to date
as the mapping of the restored index commit.

2. We should not compare the mapping version of the follower and the
leader. They are not related to one another.

Moreover, I think we should only ensure that once the restore is done,
the mapping on the follower should be at least the mapping of the copied
index commit. We don't have to sync the mapping which is updated after
we have opened a session.

Relates #36879
Closes #37887

* Change
@jimczi jimczi added v7.0.0-beta1 and removed v7.0.0 labels Feb 7, 2019
@Tim-Brooks Tim-Brooks deleted the mappings_update branch April 30, 2020 18:24
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
:Distributed Indexing/CCR Issues around the Cross Cluster State Replication features >non-issue v6.7.0 v7.0.0-beta1
Projects
None yet
Development

Successfully merging this pull request may close these issues.

5 participants