You are viewing a plain text version of this content. The canonical link for it is here.

Posted to yarn-dev@hadoop.apache.org by Krisztian Horvath <kr...@sequenceiq.com> on 2014/07/25 21:38:23 UTC

IPC ephemeral ports

Hi everyone,

I’ve added a yarn property to take control over the port binding (yarn.app.mapreduce.am.job.client.port-range) and it seems that not all the nodes care about this property as I still see logs like this on different nodes

2014-07-25 10:51:00,214 INFO [main] org.apache.hadoop.ipc.Client: Retrying connect to server: 172.17.13.2/172.17.13.2:60304. Already tried 6 time(s); retry policy is RetryUpToMaximumCountWithFixedSleep(maxRetries=50, sleepTime=1000 MILLISECONDS)

but it should be in range 40000-40050. The cluster is installed with Ambari so the config distribution shouldn’t be a problem.

Has anyone had similar issue like this?

Br,
Krisz