maedhroz commented on a change in pull request #1301:
URL: https://github.com/apache/cassandra/pull/1301#discussion_r796255662
##########
File path: src/java/org/apache/cassandra/utils/JVMStabilityInspector.java
##########
@@ -140,15 +140,26 @@ else if (t instanceof UnrecoverableIllegalStateException)
if (t instanceof FSError || t instanceof CorruptSSTableException)
isUnstable = true;
- fn.accept(t);
-
// Check for file handle exhaustion
if (t instanceof FileNotFoundException || t instanceof
FileSystemException || t instanceof SocketException)
if (t.getMessage() != null && t.getMessage().contains("Too many
open files"))
isUnstable = true;
if (isUnstable)
+ {
+ if (!StorageService.instance.isDaemonSetupCompleted())
+ FileUtils.handleStartupFSError(t);
killer.killCurrentJVM(t);
Review comment:
So if `isUnstable == true`, we're going to kill the JVM no matter what,
but is the point of calling `handleStartupFSError()` that we log a better
message when, in fact, we're not done with daemon setup? (It seems like both of
the possible consumers call it as well when the setup isn't done, so this at
least preserves the existing behavior through the reordering of this block w/
the `accept()` call...)
--
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.
To unsubscribe, e-mail: [email protected]
For queries about this service, please contact Infrastructure at:
[email protected]
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
