Hi, akshay Can you successfully execute command such as "hadoop fs -ls hdfs://987.65.43.21:8020/" on the machine you ran distcp? I think the answer is no. If both servers are reachable, I think you should check conf dfs.namenode.rpc-address or dfs.namenode.rpc-address.<nameservice>.nn1 (if HA enabled) from hdfs-site.xml to fetch the correct rpc port.
Best, Tao Yang > 在 2019年5月24日，下午3:20，akshay naidu <akshaynaid...@gmail.com> 写道： > > Hello Joey, > Just to understand distcp I am trying to copy one file. Otherwise the data is > to be copied is > 1.5TB . > Anyways I tried running -cp but looks like the issue is in connectivity. See > logs :- > hdfs dfs -cp hdfs:// > 123.45.67.89:54310/data-analytics/spike/beNginxLogs/today/123.45.67.89_2019-05-22.access.log.gz > > <http://123.45.67.89:54310/data-analytics/spike/beNginxLogs/today/123.45.67.89_2019-05-22.access.log.gz> > hdfs://987.65.43.21:50070/distCp/ > 19/05/24 07:15:22 INFO ipc.Client: Retrying connect to server: > li868-219.members.linode.com/ <http://li868-219.members.linode.com/> > 987.65.43.21:50070. Already tried 0 time(s); maxRetries=45 > 19/05/24 07:15:42 INFO ipc.Client: Retrying connect to server: > li868-219.members.linode.com/ <http://li868-219.members.linode.com/> > 987.65.43.21:50070. Already tried 1 time(s); maxRetries=45 > 19/05/24 07:16:02 INFO ipc.Client: Retrying connect to server: > li868-219.members.linode.com/ <http://li868-219.members.linode.com/> > 987.65.43.21:50070. Already tried 2 time(s); maxRetries=45 > . > . > facing same issue. > Any Idea? > Thanks. Regards > > On Fri, May 24, 2019 at 8:10 AM Joey Krabacher <jkrabac...@gmail.com > <mailto:jkrabac...@gmail.com>> wrote: > It looks like you're just trying to copy 1 file? > Why not use 'hdfs dfs -cp ...' instead? > > On Thu, May 23, 2019, 21:22 yangtao.yt <http://yangtao.yt/> > <yangtao...@alibaba-inc.com <mailto:yangtao...@alibaba-inc.com>> wrote: > Hi, akshay > > Seems it’s not distcp’s business, SocketTimeout exceptions may be caused by > network unreachable or unavailable remote server, you can communicate with > the target hdfs cluster directly on the machine you executed distcp command > to have a test. > Fully causes and suggestions given by community can be fetched from here: > https://wiki.apache.org/hadoop/SocketTimeout > <https://wiki.apache.org/hadoop/SocketTimeout> > > There is a doubt in your distcp command, why using port 50070 (http port) > instead of 8020 (rpc port) for the target hdfs cluster? I’m confusing that it > still can connect with 8020 according to your logs. > > Best, > Tao Yang > >> 在 2019年5月23日，下午8:54，akshay naidu <akshaynaid...@gmail.com >> <mailto:akshaynaid...@gmail.com>> 写道： >> >> sun.reflect.NativeConstructorAccessorImpl.newInstance0 >
smime.p7s
Description: S/MIME cryptographic signature
