honeymoose/OpenSearch
1
0
Fork 0
mirror of https://github.com/honeymoose/OpenSearch.git synced 2025-03-03 01:19:10 +00:00
Code Issues Packages Projects Releases Wiki Activity

Documentation for handling engine write failures

Browse Source
This commit is contained in:
Areek Zillur 2016-10-15 15:20:38 -04:00
parent 26f5118706
commit e195f7dd19
1 changed files with 26 additions and 17 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

43
core/src/main/java/org/elasticsearch/index/engine/InternalEngine.java
View File

@ -57,6 +57,7 @@ import org.elasticsearch.common.util.concurrent.KeyedLock;
import org.elasticsearch.common.util.concurrent.ReleasableLock;
import org.elasticsearch.index.IndexSettings;
import org.elasticsearch.index.VersionType;
import org.elasticsearch.index.engine.Engine.Operation;
import org.elasticsearch.index.mapper.Uid;
import org.elasticsearch.index.merge.MergeStats;
import org.elasticsearch.index.merge.OnGoingMerge;
@ -414,25 +415,33 @@ public class InternalEngine extends Engine {
}
/**
* When indexing a document into Lucene, Lucene distinguishes between environment related errors
* (like out of disk space) and document specific errors (like analysis chain problems) by setting
* the IndexWriter.getTragicEvent() value for the former. maybeFailEngine checks for these kind of
* errors and returns true if that is the case. We use that to indicate a document level failure
* and set the error in operation.setFailure. In case of environment related errors, the failure
* is bubbled up
* Handle failures executing write operations, distinguish persistent engine (environment) failures
* from document (request) specific failures.
* Write failures that fail the engine as a side-effect, are thrown wrapped in {@link OperationFailedEngineException}
* and document specific failures are captured through {@link Operation#setFailure(Exception)} to be handled
* at the transport level.
*/
private void handleOperationFailure(final Operation operation, final Exception e) throws OperationFailedEngineException {
private void handleOperationFailure(final Operation operation, final Exception failure) {
boolean isEnvironmentFailure;
try {
if (maybeFailEngine(operation.operationType().getLowercase(), e)) {
throw new OperationFailedEngineException(shardId,
operation.operationType().getLowercase(), operation.type(), operation.id(), e);
} else {
operation.setFailure(e);
}
// When indexing a document into Lucene, Lucene distinguishes between environment related errors
// (like out of disk space) and document specific errors (like analysis chain problems) by setting
// the IndexWriter.getTragicEvent() value for the former. maybeFailEngine checks for these kind of
// errors and returns true if that is the case. We use that to indicate a document level failure
// and set the error in operation.setFailure. In case of environment related errors, the failure
// is bubbled up
isEnvironmentFailure = (failure instanceof IllegalStateException || failure instanceof IOException)
&& maybeFailEngine(operation.operationType().getLowercase(), failure);
} catch (Exception inner) {
e.addSuppressed(inner);
throw new OperationFailedEngineException(shardId,
operation.operationType().getLowercase(), operation.type(), operation.id(), e);
// we failed checking whether the failure can fail the engine, treat it as a persistent engine failure
isEnvironmentFailure = true;
failure.addSuppressed(inner);
}
if (isEnvironmentFailure) {
throw new OperationFailedEngineException(shardId, operation.operationType().getLowercase(),
operation.type(), operation.id(), failure);
} else {
operation.setFailure(failure);
}
}
@ -540,7 +549,7 @@ public class InternalEngine extends Engine {
}
}
private long updateVersion(Engine.Operation op, long currentVersion, long expectedVersion) {
private long updateVersion(Operation op, long currentVersion, long expectedVersion) {
final long updatedVersion = op.versionType().updateVersion(currentVersion, expectedVersion);
op.updateVersion(updatedVersion);
return updatedVersion;