I have not seen the issue after changing the commit log segment size to 1024MB.
tpstats output: Pool Name Active Pending Completed Blocked All time blocked ReadStage 0 0 0 0 0 RequestResponseStage 0 0 0 0 0 MutationStage 32 40 2526143 0 0 ReadRepairStage 0 0 0 0 0 ReplicateOnWriteStage 0 0 0 0 0 GossipStage 0 0 0 0 0 AntiEntropyStage 0 0 0 0 0 MigrationStage 0 0 3 0 0 MemoryMeter 0 0 24752 0 0 MemtablePostFlusher 1 19 12939 0 0 FlushWriter 6 10 12442 1 2940 MiscStage 0 0 0 0 0 PendingRangeCalculator 0 0 1 0 0 commitlog_archiver 0 0 0 0 0 InternalResponseStage 0 0 0 0 0 HintedHandoff 0 0 0 0 0 Message type Dropped RANGE_SLICE 0 READ_REPAIR 0 PAGED_RANGE 0 BINARY 0 READ 0 MUTATION 0 _TRACE 0 REQUEST_RESPONSE 0 COUNTER_MUTATION 0 On Saturday, 19 July 2014 1:32 AM, Robert Coli <rc...@eventbrite.com> wrote: On Mon, Jul 7, 2014 at 9:30 PM, Bhaskar Singhal <bhaskarsing...@yahoo.com> wrote: I am using Cassandra 2.0.7 (with default settings and 16GB heap on quad core ubuntu server with 32gb ram) 16GB of heap will lead to significant GC pauses, and probably will not improve total performance versus 8gb heap. I continue to maintain that your "problem" is that you are writing faster than you can flush. Paste the output of "nodetool tpstats"? =Rob
