zookeeper-user mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From "Bae, Jae Hyeon" <metac...@gmail.com>
Subject Question on jute.maxbuffer property
Date Fri, 08 Nov 2013 22:55:51 GMT
Hi Zookeeper users

I think that we are having an issue similar to
https://issues.apache.org/jira/browse/ZOOKEEPER-706

I can see the following log lines repeated and I am suspecting some
application is trying to acquire a number of InterProcessMutex, one of
Apache Curator recipe.

2013-11-08 20:53:10,782 - WARN  [NIOServerCxn.Factory:
0.0.0.0/0.0.0.0:2181:NIOServerCnxn@354] - Exception causing close of
session 0xb1422ec35d9402e6 due to java.io.IOException: Len error 271603750
2013-11-08 20:53:10,783 - INFO  [NIOServerCxn.Factory:
0.0.0.0/0.0.0.0:2181:NIOServerCnxn@1001] - Closed socket connection for
client /10.45.164.229:51904 which had sessionid 0xb1422ec35d9402e6
2013-11-08 20:53:10,853 - WARN  [NIOServerCxn.Factory:
0.0.0.0/0.0.0.0:2181:NIOServerCnxn@354] - Exception causing close of
session 0xb3422ec35f3c0228 due to java.io.IOException: Len error 28936498
2013-11-08 20:53:10,854 - INFO  [NIOServerCxn.Factory:
0.0.0.0/0.0.0.0:2181:NIOServerCnxn@1001] - Closed socket connection for
client /10.154.133.50:51193 which had sessionid 0xb3422ec35f3c0228

So, the solution can be increasing jute.maxbuffer and according to the
zookeeper administrator's guide, it should be set in both server and the
client. What will happen when some clients are not setting that property? I
cannot urge all clients to setup this property and restart them. I believe
increasing the value should be fine for all existing clients but I want to
confirm nothing bad would happen.

Thank you
Best, Jae

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