I have a TransportClient connecting to a 2 node search cluster over NAT.
The client starts up fine, over time I'm experiencing a
NoNodeAvailableException though.
Are there any configuration changes recommended for clients over NAT. I've
already discovered to turn sniff off, since that hung the launch. I'm
using 0.19.8
Heres the deal, I start the app and the client connects and works fine.
After a certain amount of time, a search request fails and writes these
logs. It writes disconnect and connect log almost immediately. The next
requests fail.
2013-03-08 12:51:29,040 DEBUG (netty:94): [SearchClient] disconnected from
[[#transport#-1][inet[searchserve1/39.162.203.38:9300]]]
2013-03-08 12:51:31,993 DEBUG (netty:94): [SearchClient] connected to node
[[#transport#-1][inet[searchserve1/39.162.203.38:9300]]]
2013-03-08 12:51:47,994 DEBUG (netty:94): [SearchClient] disconnected from
[[#transport#-2][inet[searchserve2/39.162.203.39:9300]]]
2013-03-08 12:51:52,072 DEBUG (netty:94): [SearchClient] connected to node
[[#transport#-2][inet[searchserve2/39.162.203.39:9300]]]
2013-03-08 13:27:10,115 DEBUG (netty:94): [SearchClient] disconnected from
[[#transport#-2][inet[searchserve2/39.162.203.39:9300]]]
2013-03-08 13:27:14,272 DEBUG (netty:94): [SearchClient] connected to node
[[#transport#-2][inet[searchserve2/39.162.203.39:9300]]]
2013-03-08 13:27:20,975 DEBUG (netty:94): [SearchClient] disconnected from
[[#transport#-1][inet[searchserve1/39.162.203.38:9300]]]
2013-03-08 13:27:24,303 DEBUG (netty:94): [SearchClient] connected to node
[[#transport#-1][inet[searchserve1/39.162.203.38:9300]]]
Apache, Apache Lucene, Apache Hadoop, Hadoop, HDFS and the yellow elephant
logo are trademarks of the
Apache Software Foundation
in the United States and/or other countries.