Just wanted to check on the operation of nodetool repair. I'm running the trunk nightly from 2010-08-26_13-52-57
I ran it this morning against one node in a 4 node cluster, on each node in the cluster I saw this in the log
INFO [manual-repair-6516fb70-d654-4fe2-815b-5d3a6b77106d] 2010-09-01 09:45:13,891 AntiEntropyService.java (line 803) Waiting for repair requests to: []
In the comments for https://issues.apache.org/jira/browse/CASSANDRA-1432 Stu said that this line indicates the cluster thinks the RF is 1, however I've double checked using the cassandra-cli and it's 3 for this keyspace.
I turned logging up to DEBUG and these look like the messages associated with running the repair...
DEBUG [Timer-1] 2010-09-01 10:06:56,660 LoadDisseminator.java (line 37) Disseminating load info ...
DEBUG [manual-repair-278ebd9f-21c0-43b2-94c2-9c707518c30c] 2010-09-01 10:07:18,896 StorageService.java (line 1251) computing ranges for 42535295865117307932921825928971026432, 8507059173023461586584365
1857942052864, 127605887595351923798765477786913079296, 170141183460469231731687303715884105728
INFO [manual-repair-278ebd9f-21c0-43b2-94c2-9c707518c30c] 2010-09-01 10:07:18,897 AntiEntropyService.java (line 803) Waiting for repair requests to: []
DEBUG [Timer-1] 2010-09-01 10:07:56,661 LoadDisseminator.java (line 37) Disseminating load info ...
According to code in AntiEntrophyService around line 804 I should see another INFO message is the repair requests were sent successfully. After 10 minutes there were no messages in the log from either that thread or the AntiEntrophyService logger.
Is there anything else I can do to check that AES is working or is the log message really saying that there is nothing to do?
Thanks
Aaron
