Support shard promotion with Segment Replication.

[mch2]

Signed-off-by: Marc Handalian [email protected]

Description

This change adds basic failover support with segment replication. Once selected, a replica will commit its SegmentInfos and reopen a writeable engine. The replica will also remove all other commits so that this commit is selected when the writeable engine is opened. It is possible that this commit is not considered 'safe' by the primary, meaning its max seqNo is higher than the global cp. While an edge case, we never want replicas to reindex with segment replication enabled, so if the global cp has not been updated yet we do not want to revert to a safe commit. This change also updates how SegmentReplicationCheckpointPublisher is wired up within IndexShard so that once promoted the new primary can publish checkpoints.

This PR does not handle edge cases of promotion while there are ongoing replication events, that will be covered in a separate issue.

Issues Resolved

closes #3989

Check List

• New functionality includes testing.
  • All tests pass
• New functionality has been documented.
  • New functionality has javadoc added
• Commits are signed per the DCO using --signoff

By submitting this pull request, I confirm that my contribution is made under the terms of the Apache 2.0 license.
For more information on following Developer Certificate of Origin and signing off your commits, please check here.

[github-actions]

Gradle Check (Jenkins) Run Completed with:

• RESULT: SUCCESS ✅
• URL: https://build.ci.opensearch.org/job/gradle-check/1468/
• CommitID: c28905a

[codecov-commenter]

Codecov Report

Merging #4135 (28fea5f) into main (237f1a5) will decrease coverage by 0.00%.
The diff coverage is 76.92%.

@@             Coverage Diff              @@
##               main    #4135      +/-   ##
============================================
- Coverage     70.65%   70.64%   -0.01%     
- Complexity    57075    57145      +70     
============================================
  Files          4606     4606              
  Lines        274706   274737      +31     
  Branches      40228    40231       +3     
============================================
- Hits         194103   194099       -4     
- Misses        64280    64374      +94     
+ Partials      16323    16264      -59     

Impacted Files	Coverage Δ
...c/main/java/org/opensearch/index/IndexService.java	73.86% <0.00%> (-0.23%)	⬇️
...nsearch/index/shard/CheckpointRefreshListener.java	88.88% <0.00%> (-11.12%)	⬇️
...in/java/org/opensearch/index/shard/IndexShard.java	69.07% <73.33%> (-0.58%)	⬇️
...rc/main/java/org/opensearch/index/store/Store.java	81.30% <81.25%> (-0.60%)	⬇️
.../opensearch/index/engine/NRTReplicationEngine.java	76.92% <100.00%> (+1.52%)	⬆️
...ation/OpenSearchIndexLevelReplicationTestCase.java	89.81% <100.00%> (+0.02%)	⬆️
...java/org/opensearch/client/indices/DataStream.java	0.00% <0.00%> (-76.09%)	⬇️
...n/indices/forcemerge/ForceMergeRequestBuilder.java	0.00% <0.00%> (-75.00%)	⬇️
...adonly/AddIndexBlockClusterStateUpdateRequest.java	0.00% <0.00%> (-75.00%)	⬇️
.../opensearch/client/indices/CloseIndexResponse.java	17.50% <0.00%> (-60.00%)	⬇️
... and 499 more

Help us with your feedback. Take ten seconds to tell us how you rate us. Have a feature suggestion? Share it here.

[mch2]

I've added a commit to this ensuring cancelling primary allocation succeeds and that the replica is promoted & primary recreated as a replica. In testing that I found we were failing to publish a replication checkpoint if the primary flushed during close. That is now fixed, the shard must be open for us to publish the replication cp.

[github-actions]

Gradle Check (Jenkins) Run Completed with:

• RESULT: UNSTABLE ❌
• URL: https://build.ci.opensearch.org/job/gradle-check/1597/
• CommitID: 8eedb4a

[github-actions]

Gradle Check (Jenkins) Run Completed with:

• RESULT: SUCCESS ✅
• URL: https://build.ci.opensearch.org/job/gradle-check/1612/
• CommitID: 8eedb4a

[github-actions]

Gradle Check (Jenkins) Run Completed with:

• RESULT: SUCCESS ✅
• URL: https://build.ci.opensearch.org/job/gradle-check/1755/
• CommitID: de626ff10fc1ad4780b5bae593cf45b55b3112f2

[github-actions]

Gradle Check (Jenkins) Run Completed with:

• RESULT: FAILURE ❌
• URL: https://build.ci.opensearch.org/job/gradle-check/1811/
• CommitID: 49c483851d8c785bb000924f44ebe99d95187738

[github-actions]

Gradle Check (Jenkins) Run Completed with:

• RESULT: FAILURE ❌
• URL: https://build.ci.opensearch.org/job/gradle-check/1812/
• CommitID: d7b3241

[github-actions]

Gradle Check (Jenkins) Run Completed with:

• RESULT: SUCCESS ✅
• URL: https://build.ci.opensearch.org/job/gradle-check/1813/
• CommitID: d7b3241

[github-actions]

Gradle Check (Jenkins) Run Completed with:

• RESULT: UNSTABLE ❌
• URL: https://build.ci.opensearch.org/job/gradle-check/1816/
• CommitID: 28fea5f

[github-actions]

Gradle Check (Jenkins) Run Completed with:

• RESULT: SUCCESS ✅
• URL: https://build.ci.opensearch.org/job/gradle-check/1818/
• CommitID: 28fea5f

[opensearch-trigger-bot]

The backport to 2.x failed:

The process '/usr/bin/git' failed with exit code 1

To backport manually, run these commands in your terminal:

# Fetch latest updates from GitHub
git fetch
# Create a new working tree
git worktree add .worktrees/backport-2.x 2.x
# Navigate to the new working tree
cd .worktrees/backport-2.x
# Create a new branch
git switch --create backport/backport-4135-to-2.x
# Cherry-pick the merged commit of this pull request and resolve the conflicts
git cherry-pick -x --mainline 1 f65e02d1b910bd0a1990868bfa5d12ba829bbbd5
# Push it to GitHub
git push --set-upstream origin backport/backport-4135-to-2.x
# Go back to the original working tree
cd ../..
# Delete the working tree
git worktree remove .worktrees/backport-2.x

Then, create a pull request where the base branch is 2.x and the compare/head branch is backport/backport-4135-to-2.x.