Add test for PutFollowAction on a closed index
#38236
Merged
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
This is related to elastic#35975. Currently when an index falls behind a leader it encounters a fatal exception. This commit adds a test for that scenario. Additionally, it tests that the user can stop following, close the follower index, and put follow again. After the indexing is re-bootstrapped, it will recover the documents it lost in normal following operations.
Tim-Brooks
added
>test
|
Pinging @elastic/es-distributed |
32 tasks
jasontedor
approved these changes
Feb 2, 2019
martijnvg
approved these changes
Feb 2, 2019
| .filter(Objects::nonNull) | ||
| .map(ExceptionsHelper::unwrapCause) | ||
| .filter(e -> e instanceof ResourceNotFoundException) | ||
| .map(e -> (ResourceNotFoundException) e) |
There was a problem hiding this comment.
Maybe also exclude exception that do not have es.requested_operations_missing as metadata? Then we are even more sure that the exception is the right one.
dnhatn
approved these changes
Feb 2, 2019
| for (int i = 0; i < numDocs; i++) { | ||
| final String source = String.format(Locale.ROOT, "{\"f\":%d}", i * 2); | ||
| leaderClient().prepareIndex("index1", "doc", Integer.toString(i)).setSource(source, XContentType.JSON).get(); | ||
| leaderClient().admin().indices().flush(new FlushRequest("index1").force(true)).actionGet(); |
There was a problem hiding this comment.
nit: one flush before the force_merge is good enough :).
| assertTrue(response.isFollowIndexShardsAcked()); | ||
| assertTrue(response.isIndexFollowingStarted()); | ||
|
|
||
| final Map<ShardId, Long> firstBatchNumDocsPerShard = new HashMap<>(); |
There was a problem hiding this comment.
nit: maybe use "assertIndexFullyReplicatedToFollower"?
| assertTrue(response2.isFollowIndexShardsAcked()); | ||
| assertTrue(response2.isIndexFollowingStarted()); | ||
|
|
||
| final Map<ShardId, Long> secondBatchNumDocsPerShard = new HashMap<>(); |
There was a problem hiding this comment.
nit: maybe use "assertIndexFullyReplicatedToFollower"?
ywelsch
suggested changes
Feb 4, 2019
| assertThat(exceptions.size(), greaterThan(0)); | ||
| }); | ||
|
|
||
| pauseFollow("index2"); |
There was a problem hiding this comment.
what happens if we close before pausing following? What happens if we don't explicitly pause following at all? Let's also check these scenarios.
There was a problem hiding this comment.
what happens if we close before pausing following?
It works. We allow closing a follower index. There is a test that checks this behavior testCloseFollowIndex.
Let's also check these scenarios.
I added a test for the failure scenarios in the restart following case (index still open and following not paused).
ywelsch
approved these changes
Feb 4, 2019
Tim-Brooks
added a commit
to Tim-Brooks/elasticsearch
that referenced
this pull request
Feb 4, 2019
This is related to elastic#35975. Currently when an index falls behind a leader it encounters a fatal exception. This commit adds a test for that scenario. Additionally, it tests that the user can stop following, close the follower index, and put follow again. After the indexing is re-bootstrapped, it will recover the documents it lost in normal following operations.
Tim-Brooks
added a commit
that referenced
this pull request
Feb 5, 2019
This is related to #35975. Currently when an index falls behind a leader it encounters a fatal exception. This commit adds a test for that scenario. Additionally, it tests that the user can stop following, close the follower index, and put follow again. After the indexing is re-bootstrapped, it will recover the documents it lost in normal following operations.
jasontedor
added a commit
to jasontedor/elasticsearch
that referenced
this pull request
Feb 5, 2019
…-lease-expiration * elastic/master: (24 commits) Add support for API keys to access Elasticsearch (elastic#38291) Add typless client side GetIndexRequest calls and response class (elastic#37778) Limit token expiry to 1 hour maximum (elastic#38244) add docs saying mixed-cluster ILM is not supported (elastic#37954) Skip unsupported languages for tests (elastic#38328) Deprecate `_type` in simulate pipeline requests (elastic#37949) Mute testCannotShrinkLeaderIndex (elastic#38374) Tighten mapping syncing in ccr remote restore (elastic#38071) Add test for `PutFollowAction` on a closed index (elastic#38236) Fix SSLContext pinning to TLSV1.2 in reload tests (elastic#38341) Mute RareClusterStateIT.testDelayedMappingPropagationOnReplica (elastic#38357) Deprecate types in rollover index API (elastic#38039) Types removal - fix FullClusterRestartIT warning expectations (elastic#38310) Fix ILM explain response to allow unknown fields (elastic#38054) Mute testFollowIndexAndCloseNode (elastic#38360) Docs: Drop inline callout from scroll example (elastic#38340) Deprecate HLRC security methods (elastic#37883) Remove types from Monitoring plugin "backend" code (elastic#37745) Add Composite to AggregationBuilders (elastic#38207) Clarify slow cluster-state log messages (elastic#38302) ...
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.
This is related to #35975. Currently when an index falls behind a leader
it encounters a fatal exception. This commit adds a test for that
scenario. Additionally, it tests that the user can stop following, close
the follower index, and put follow again. After the indexing is
re-bootstrapped, it will recover the documents it lost in normal
following operations.