One thought I had would be to grep for all lines in the logs for the equivalent of
grep "Cleaning up SSTableReader(path='/data/cassandra/keyspace/cf/" /var/log/cassandra/system.log | wc -l which would give you the number of sstables processed for that column family. Then you could compare that to the number of total sstables in your data dir for that cf for a rough percentage of sstables that have been processed. Thoughts? On 11/29/12 10:36 AM, "Michael Kjellman" <mkjell...@barracuda.com> wrote: >compactionstats does not show the overall progress it has made on cleanup >operations for an entire keyspace -- It just shows the current progress on >each sstable it is working on. How would you suggest you abstract those >stats to guesstimate overall cleanup progress across all sstables in that >keyspace? > >On 11/29/12 10:20 AM, "Jeremy Hanna" <jeremy.hanna1...@gmail.com> wrote: > >>You can do check nodetool compactionstats to see progress for current >>cleanup operations. It essentially traverses all of your sstables and >>removes data that the node isn't responsible for. So that's the overall >>operation, so you would estimate in terms of how long it would take to go >>through all of those. >> >>On Nov 29, 2012, at 10:37 AM, Michael Kjellman <mkjell...@barracuda.com> >>wrote: >> >>> I know there is very little insight currently into long running JMX >>>operations but has anyone figured out any way clever to determine >>>approximate progress on cleanup operations? Merci. >>> >>> Best, >>> Michael >>> >>> ---------------------------------- >>> 'Like' us on Facebook for exclusive content and other resources on all >>>Barracuda Networks solutions. >>> Visit http://barracudanetworks.com/facebook >>> >> > > >'Like' us on Facebook for exclusive content and other resources on all >Barracuda Networks solutions. >Visit http://barracudanetworks.com/facebook > > 'Like' us on Facebook for exclusive content and other resources on all Barracuda Networks solutions. Visit http://barracudanetworks.com/facebook