Details
-
Bug
-
Resolution: Fixed
-
Major
-
None
-
None
-
Security Level: Public
-
None
Description
When running the failover integration tests, it is possible that a node is removed from the cluster, but operations try to go towards the node. This results either in Assertion Exceptions or NPEs down the stack when trying to authenticate (but its not anymore part of the cluster).
For example:
69.15 SDKD: Jan 25, 2013 6:40:23 AM com.couchbase.client.CouchbaseConnection reconfigure
69.15 SDKD: INFO: Scheduling Node localhost-27.local/10.3.121.207:11210for shutdown.
leads into:
75.15 SDKD: Exception in thread "Memcached IO over
" java.lang.AssertionError: Couldn't find node connected to localhost-27.local/10.3.121.207:11210
75.15 SDKD: at net.spy.memcached.MemcachedClient.findNode(MemcachedClient.java:2117)
75.15 SDKD: at net.spy.memcached.MemcachedClient.connectionEstablished(MemcachedClient.java:2106)
75.15 SDKD: at net.spy.memcached.MemcachedConnection.connected(MemcachedConnection.java:352)
75.15 SDKD: at net.spy.memcached.MemcachedConnection.handleIO(MemcachedConnection.java:428)
75.15 SDKD: at net.spy.memcached.MemcachedConnection.handleIO(MemcachedConnection.java:247)
75.15 SDKD: at com.couchbase.client.CouchbaseConnection.run(CouchbaseConnection.java:266)
Attachments
Issue Links
- duplicates
-
JCBC-70 Client fails to reconnect to server of non-default memcached bucket after failover and add back
- Resolved