http://www.datastax.com/support-forums/topic/how-to-clean-cassandra-hints-files Software, Support, and Training for Apache Cassandra en-US Thu, 23 May 2013 08:35:52 +0000 http://bbpress.org/?v=1.0.3 q http://www.datastax.com/support-forums/search.php http://www.datastax.com/support-forums/topic/how-to-clean-cassandra-hints-files#post-6610 Fri, 21 Sep 2012 09:58:41 +0000 xedin 6610@http://www.datastax.com/support-forums/ <p>all the state changes are handled by the periodic Gossiper thread or FailureDetector so such log messages are just coincidence with nodetool ring which was running simultaneously with nodetool. But if you see UP state after you run ring and log had notified that it's DEAD that is a bug Gossiper and you can report it to <a href="https://issues.apache.org/jira/browse/CASSANDRA" rel="nofollow">https://issues.apache.org/jira/browse/CASSANDRA</a> to be dealt with. </p> http://www.datastax.com/support-forums/topic/how-to-clean-cassandra-hints-files#post-6606 Fri, 21 Sep 2012 06:46:03 +0000 ouzheyi 6606@http://www.datastax.com/support-forums/ <p>We found that all nodes's status are Up when we used the nodetool ring command while "system.log.6: INFO [GossipTasks:1] 2012-09-13 17:09:32,966 Gossiper.java (line 818) InetAddress /130.51.11.223 is now dead." appeared. Why ? And we want to know how to reduce this. </p> http://www.datastax.com/support-forums/topic/how-to-clean-cassandra-hints-files#post-6593 Thu, 20 Sep 2012 10:37:29 +0000 xedin 6593@http://www.datastax.com/support-forums/ <p>Alright, you can go on just deleting the files then. </p> http://www.datastax.com/support-forums/topic/how-to-clean-cassandra-hints-files#post-6592 Thu, 20 Sep 2012 10:33:56 +0000 ouzheyi 6592@http://www.datastax.com/support-forums/ <p>hi<br /> These three nodes are our data nodes, we just want to clean files like HintsColumnFamily-* without affecting the cluster </p> http://www.datastax.com/support-forums/topic/how-to-clean-cassandra-hints-files#post-6591 Thu, 20 Sep 2012 09:07:16 +0000 xedin 6591@http://www.datastax.com/support-forums/ <p>Hi, are you ok with whatever happened three of your nodes? If so then you can just decommission those nodes from the cluster using nodetool and safely delete hits from Cassandra's system data directory, which could be usually find in - /var/lib/cassandra/data/system/HintsColumnFamily-* </p> http://www.datastax.com/support-forums/topic/how-to-clean-cassandra-hints-files#post-6589 Thu, 20 Sep 2012 06:21:01 +0000 ouzheyi 6589@http://www.datastax.com/support-forums/ <p>hi all </p> <p>Our cluster uses cassandra 1.0.8 with replication = 2. We found that many HintsColumnFamily files exist in the system dir, even up to n*100G. Then i check the logs,found that :</p> <p>system.log.6: INFO [GossipTasks:1] 2012-09-13 17:09:32,966 Gossiper.java (line 818) InetAddress /130.51.11.223 is now dead.<br /> system.log.6: INFO [GossipTasks:1] 2012-09-13 17:15:01,061 Gossiper.java (line 818) InetAddress /130.51.11.220 is now dead.<br /> system.log.6: INFO [GossipTasks:1] 2012-09-13 17:18:27,124 Gossiper.java (line 818) InetAddress /130.51.11.221 is now dead.</p> <p>We had runned repair system HintsColumnFamily command, but hints files still exist.</p> <p>If you can give some suggestions, we may thank u very much ! </p>