zookeeper-flush WAL file slowly

2014-05-15 09:08:01,803 [myid:5] - WARN  [SyncThread:5:FileTxnLog@321] - fsync-ing the write ahead log in SyncThread:5 took 15154ms which will adversely effect operation latency. See the ZooKeeper troubleshooting guide

 TODO

-solution A:decrease high io

 

 

-solution B:decrease nodes registered to zk so also indirectly decrease edits written to WAL

 

-C :check GC

-D move zk ensemble to io-light nodes