On Wed, Mar 18, 2015 at 10:14 AM, Caraballo, Rafael < [email protected]> wrote:
> After upgrading a 3 node Cassandra cluster from 1.2.19 to 2.0.12, I have > an event storm of “ SliceQueryFilter” messages flooding the Cassandra > system.log file. > > > > How can I stop this event storm? > > > As the message says : " (see tombstone_warn_threshold) " The thing you are being warned about is that your write pattern results in a significant number of tombstones. In general this is a smell of badness in Cassandra, which is why the log message exists. To Resolve : 1) Increase tombstone_warn_threshold OR 2) Stop creating so many tombstones =Rob http://twitter.com/rcolidba
