jbarefoot opened a new pull request #699: Fix so dfs.client.failover.max.attempts is respected correctly URL: https://github.com/apache/hadoop/pull/699 Without this change, you would get incorrect behavior in that you would always have to set `dfs.client.failover.max.attempts` to be +1 greater to have the desired behavior, e.g. if you want it to attempt failover exactly once, you would have to set `dfs.client.failover.max.attempts=2`. Without this change, if you set `dfs.client.failover.max.attempts=1`, to attempt to failover just once time, instead it wouldn't try at all, and you see this log message: ``` org.apache.hadoop.io.retry.RetryInvocationHandler - Exception while invoking class org.apache.hadoop.hdfs.protocolPB.ClientNamenodeProtocolTranslatorPB.getFileInfo over hadoop-node-01.docker.infra.atscale.com/127.0.0.21:8020. Not retrying because failovers (1) exceeded maximum allowed (1) ``` Note that the non-failover retries just below this change is correct.
---------------------------------------------------------------- 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 --------------------------------------------------------------------- To unsubscribe, e-mail: [email protected] For additional commands, e-mail: [email protected]
