http://www.datastax.com/support-forums/topic/ttl-and-repair Software, Support, and Training for Apache Cassandra en-US Sun, 26 May 2013 08:12:25 +0000 http://bbpress.org/?v=1.0.3 q http://www.datastax.com/support-forums/search.php http://www.datastax.com/support-forums/topic/ttl-and-repair#post-1198 Wed, 22 Feb 2012 06:56:09 +0000 Anonymous 1198@http://www.datastax.com/support-forums/ <p>Thank you for your answer,<br /> We have a rather large dataset with some CF's that are close to 100GB, and have not yet run a repair on this node. We tried to <a href="http://www.gold4fun.com/" title="RuneScape Gold">runescape gold</a> run repair, but after a week we stopped because performance was impacted.<br /> Can you tell me <a href="http://www.gold4fun.com/Runescape-accounts.html" title="runescape accounts">runescape accounts</a> what would be the consequence of not repairing?<br /> And if there is something that could speed up the repair?<br /> thanks! </p> http://www.datastax.com/support-forums/topic/ttl-and-repair#post-1186 Tue, 21 Feb 2012 04:52:10 +0000 rbranson 1186@http://www.datastax.com/support-forums/ <p>There is always going to be a cost in running the anti-entropy system (repair). It's fundamentally a complex operation. If the other mechanisms (hinted handoff and read repair) do not repair data after replication failure, the consequence is that data can become synchronized between replicas. At best, this means stale data will be read for low consistency operations. At worst, it means deleted data will reappear when tombstones are cleaned up. </p> http://www.datastax.com/support-forums/topic/ttl-and-repair#post-1184 Mon, 20 Feb 2012 13:26:38 +0000 Anonymous 1184@http://www.datastax.com/support-forums/ <p>Thank you for your answer,</p> <p>We have a rather large dataset with some CF's that are close to 100GB, and have not yet run a repair on this node.<br /> We tried to run repair, but after a week we stopped because performance was impacted.</p> <p>Can you tell me what would be the consequence of not repairing?<br /> And if there is something that could speed up the repair?</p> <p>thanks! </p> http://www.datastax.com/support-forums/topic/ttl-and-repair#post-1177 Sat, 18 Feb 2012 01:01:46 +0000 rbranson 1177@http://www.datastax.com/support-forums/ <p>The compaction process simply converts expired columns into tombstones, so you will still need to run a regular repair. </p> http://www.datastax.com/support-forums/topic/ttl-and-repair#post-1172 Fri, 17 Feb 2012 09:04:16 +0000 Anonymous 1172@http://www.datastax.com/support-forums/ <p>Hi,</p> <p>I have a rather simple question for which we can't find a clear answer:</p> <p>We insert all our data with a TTL of 1 month.<br /> Do we still need to run manual regular repairs or will CG get rid of the data automatically?</p> <p>thanks! </p>