saintstack commented on a change in pull request #956: HBASE-23601:
OutputSink.WriterThread exception gets stuck and repeated indefinietly
URL: https://github.com/apache/hbase/pull/956#discussion_r360489711
##########
File path:
hbase-server/src/main/java/org/apache/hadoop/hbase/wal/WALSplitter.java
##########
@@ -478,6 +478,10 @@ void checkForErrors() throws IOException {
throw new RuntimeException(thrown);
}
}
+
+ public Throwable clearError(){
+ return thrown.getAndSet(null);
+ }
Review comment:
I see we make the pipelinecontroller once and use it thereafter -- it goes
down into the WriterThread.... . I was going to suggest just create a new
Controller on each write but looks like too much of a refactor.
Would it be cleaner if checkForErrors cleared any errors found... ; i.e.
checkForErrors throws but before it throws it clears any exception in
Controller?
----------------------------------------------------------------
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.
For queries about this service, please contact Infrastructure at:
[email protected]
With regards,
Apache Git Services
