jeffsaremi commented on issue #1480: HDFS-14857. FS operations fail in HA mode: DataNode fails to connect to NameNode URL: https://github.com/apache/hadoop/pull/1480#issuecomment-539308184 I went back and revised some of the code. It looked like we're trying to do some really important stuff right inside the perfromFailover() method. This didn't feel right and we didn't want to deal with all sorts of things that could go wrong. Instead we're doing the minimal here -- making sure the proxy is nullified -- and then when the Retry loop calls the getProxy() we deal with actual IP resetting. This was important since we already had a lazy initializer inside AbstractNNFailoverProxyProvider and we could just reuse that . Also there seemed to be some overlap in the old way of resettingProxy and what AbstractNNFailoverProxyProvider is doing. So we're avoiding too much redundancy. Inside AbstractNNFailoverProxyProvider there are actually two spots that this could happen: createProxyIfNeeded() and getProxyAddresses() either one could be altered to handle the IP address reset as well.
---------------------------------------------------------------- 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]
