hadoop-yarn-dev mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From Krisztian Horvath <krisztian.horv...@sequenceiq.com>
Subject IPC ephemeral ports
Date Fri, 25 Jul 2014 19:38:23 GMT
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: 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?


  • Unnamed multipart/alternative (inline, None, 0 bytes)
View raw message