running 2 nodes wso2 EI 6.4.0 behind a load balancer (round robbin) for months
now, suddenly yesterday one node was very very busy without any good reason..
In the logs we found thousands and thousands of this message :
[2019-09-30 16:41:06,440] [-1] [] [PassThrough HTTPS-Listener I/O dispatcher
Listener] WARN
{org.apache.synapse.transport.passthru.PassThroughHttpSSLListener} - System
may be unstable: HTTPS ListeningIOReactor encountered a checked exception : Too
many open files
java.io.IOException: Too many open files
at sun.nio.ch.ServerSocketChannelImpl.accept0(Native Method)
at
sun.nio.ch.ServerSocketChannelImpl.accept(ServerSocketChannelImpl.java:422)
at
sun.nio.ch.ServerSocketChannelImpl.accept(ServerSocketChannelImpl.java:250)
at
org.apache.http.impl.nio.reactor.DefaultListeningIOReactor.processEvent(DefaultListeningIOReactor.java:170)
at
org.apache.http.impl.nio.reactor.DefaultListeningIOReactor.processEvents(DefaultListeningIOReactor.java:153)
at
org.apache.http.impl.nio.reactor.AbstractMultiworkerIOReactor.execute(AbstractMultiworkerIOReactor.java:349)
at
org.apache.synapse.transport.passthru.core.PassThroughListeningIOReactorManager$1.run(PassThroughListeningIOReactorManager.java:536)
at java.lang.Thread.run(Thread.java:748)
We needed to stop then restart the process.
This is strange since on this linux machine we configured ulimit to "unlimited"
and no particular intensive traffic was found at this moment.
We now are afraid this can be come again, what could be the origin of this
problem ? any idea to avoid it ?
Thanks
Bernard
_______________________________________________
Dev mailing list
[email protected]
http://wso2.org/cgi-bin/mailman/listinfo/dev
