throttle your writes when they start timing out (moved to user@, bcc dev@)
On Thu, Jul 1, 2010 at 9:08 PM, Peng Guo <gpcus...@gmail.com> wrote: > Hi > > I am doing a test which 30 nodes. > > After I start 300 data insert process, I looked at the tpstats: > > [ dw-greenplum-4] Pool Name Active Pending > Completed > [ dw-greenplum-4] FILEUTILS-DELETE-POOL 0 0 > 14 > [ dw-greenplum-4] STREAM-STAGE 0 0 > 0 > [ dw-greenplum-4] RESPONSE-STAGE 0 0 > 24 > [ dw-greenplum-4] ROW-READ-STAGE 0 0 > 0 > [ dw-greenplum-4] MESSAGE-DESERIALIZER-POOL 512 709997 > 3829734 > [ dw-greenplum-4] GMFD 0 0 > 1678 > [ dw-greenplum-4] CONSISTENCY-MANAGER 0 0 > 0 > [ dw-greenplum-4] ROW-MUTATION-STAGE 512 4606 > 4350906 > [ dw-greenplum-4] MESSAGE-STREAMING-POOL 0 0 > 0 > [ dw-greenplum-4] LOAD-BALANCER-STAGE 0 0 > 0 > [ dw-greenplum-4] FLUSH-SORTER-POOL 0 0 > 0 > [ dw-greenplum-4] MEMTABLE-POST-FLUSHER 0 0 > 134 > [ dw-greenplum-4] FLUSH-WRITER-POOL 0 0 > 134 > [ dw-greenplum-4] AE-SERVICE-STAGE 0 0 > 1 > [ dw-greenplum-4] HINTED-HANDOFF-POOL 1 27 > 0 > > How could I solve this problem? > > Thx. > > -- > Regards > Peng Guo > -- Jonathan Ellis Project Chair, Apache Cassandra co-founder of Riptano, the source for professional Cassandra support http://riptano.com