[Neo-report] [Bug #1333] Client nodes are not cleared in the node manager when the cluster fall back from service to verification state
Grégory Wisniewski
gregory at nexedi.com
Thu Oct 14 10:48:41 CEST 2010
Bug : Client nodes are not cleared in the node manager when the cluster fall back from service to verification state
Status : Open
Date : 2009/10/16
Link : https://www.tiolive.com/nexedi/bug_module/1333/view
Reporter : Grégory Wisniewski
Request Project : NEO R&D
Description:
Here a scenario:
- Start a cluster with one storage
- Start a client, check with neoctl that it's well known
- Cut the storage node, the cluster is no longer operational and fall back to verification stage
- Ask neoctl the client node list, the client is still here
- Restart the storage node, the cluster become operational
- Display again the client list, two nodes are known, the old (unconnected) one is still here.
This can cause a minor memory leak.
Messages :
++++++ Message #3 submitted by Grégory Wisniewski on 2010/10/14 08:48:41 UTC ++++++
Tested in r2354.
++++++ Message #2 submitted by Grégory Wisniewski on 2010/10/14 08:26:16 UTC ++++++
Fixed by r1425.
++++++ Message #1 submitted by Grégory Wisniewski on 2009/10/16 14:34:13.691000 GMT+2 ++++++
More information about the Neo-report
mailing list