StreamCorruptedException after upgrading to 0.20.0RC1

Elastic Stack Elasticsearch
1

Upgraded a 12 node cluster from 0.19.2 to 0.20.0RC1. Plugins (head/bigdesk)
were reinstalled from scratch, but the service wrapper and config
directories are identical from the previous version. The cluster contained
two indices built with 0.19.2

No clients are connected (except for the plugins). Reallocation was not
disabled before the restart, but the nodes were all stopped and started
within seconds of each other. Cluster was originally yellow since the
shards were initializing, but it eventually went to green.

Logs are quickly filling up with these exceptions:

[2012-11-05 09:06:03,419][WARN ][transport.netty ] [srch-lv108]
exception caught on transport layer [[id: 0xb84d3508,
/192.168.52.150:40744:> /192.168.52.158:9300]],
closing connection
java.io.StreamCorruptedException: invalid internal transport message format
at
org.elasticsearch.transport.netty.SizeHeaderFrameDecoder.decode(SizeHeaderFrameDecoder.java:27)
at
org.elasticsearch.common.netty.handler.codec.frame.FrameDecoder.callDecode(FrameDecoder.java:422)
at
org.elasticsearch.common.netty.handler.codec.frame.FrameDecoder.cleanup(FrameDecoder.java:478)
at
org.elasticsearch.common.netty.handler.codec.frame.FrameDecoder.channelDisconnected(FrameDecoder.java:366)
...

Cheers,

Ivan

--

2

Maybe I should pay more attention to the error messages. The exception
listed the servers in question ans showed that one rogue client was still
up.

Back to testing 0.20 ...

On Mon, Nov 5, 2012 at 9:08 AM, Ivan Brusic ivan@brusic.com wrote:

Upgraded a 12 node cluster from 0.19.2 to 0.20.0RC1. Plugins
(head/bigdesk) were reinstalled from scratch, but the service wrapper and
config directories are identical from the previous version. The cluster
contained two indices built with 0.19.2

No clients are connected (except for the plugins). Reallocation was not
disabled before the restart, but the nodes were all stopped and started
within seconds of each other. Cluster was originally yellow since the
shards were initializing, but it eventually went to green.

Logs are quickly filling up with these exceptions:

[2012-11-05 09:06:03,419][WARN ][transport.netty ] [srch-lv108]
exception caught on transport layer [[id: 0xb84d3508, /
192.168.52.150:40744 :> /192.168.52.158:9300]], closing connection
java.io.StreamCorruptedException: invalid internal transport message format
at
org.elasticsearch.transport.netty.SizeHeaderFrameDecoder.decode(SizeHeaderFrameDecoder.java:27)
at
org.elasticsearch.common.netty.handler.codec.frame.FrameDecoder.callDecode(FrameDecoder.java:422)
at
org.elasticsearch.common.netty.handler.codec.frame.FrameDecoder.cleanup(FrameDecoder.java:478)
at
org.elasticsearch.common.netty.handler.codec.frame.FrameDecoder.channelDisconnected(FrameDecoder.java:366)
...

Cheers,

Ivan

--