rmatharu commented on a change in pull request #1104: SAMZA-2266: Introduce a
backoff when there are repeated failures for host-affinity allocations
URL: https://github.com/apache/samza/pull/1104#discussion_r303530700
##########
File path:
samza-core/src/main/java/org/apache/samza/clustermanager/ContainerProcessManager.java
##########
@@ -208,32 +210,45 @@ private static AbstractContainerAllocator
buildContainerAllocator(boolean hostAf
}
public boolean shouldShutdown() {
- log.debug("ContainerProcessManager state: Completed containers: {},
Configured containers: {}, Are there too many failed containers: {}, Is
allocator thread alive: {}",
+ LOG.debug("ContainerProcessManager state: Completed containers: {},
Configured containers: {}, Are there too many failed containers: {}, Is
allocator thread alive: {}",
state.completedProcessors.get(), state.processorCount,
tooManyFailedContainers ? "yes" : "no", allocatorThread.isAlive() ? "yes" :
"no");
if (exceptionOccurred != null) {
- log.error("Exception in container process manager", exceptionOccurred);
+ LOG.error("Exception in container process manager", exceptionOccurred);
throw new SamzaException(exceptionOccurred);
}
return tooManyFailedContainers || state.completedProcessors.get() ==
state.processorCount.get() || !allocatorThread.isAlive();
}
public void start() {
- log.info("Starting the container process manager");
+ LOG.info("Starting the container process manager");
+
+ int containerRetryCount = clusterManagerConfig.getContainerRetryCount();
Review comment:
Could this entire config handling logic (224-237) be moved to a method in
ClusterManagerConfig?
----------------------------------------------------------------
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
