Hi, Is your cluster healthy? Try running ksck and see if it shows all good: http://kudu.apache.org/docs/command_line_tools_reference.html#cluster-ksck
There's nothing in those client logs that would indicate that you need to adjust --flush_threshold_mb, in fact that's almost no reason to change this configuration. See a full tablet server log could be helpful too. J-D On Thu, Jul 20, 2017 at 4:33 AM, Guangchao Deng <[email protected]> wrote: > hi, > I get the erro when i insert records into a table(20 columns, 50 > partitions with hash): > > Timed out: Failed to write batch of 699 ops to tablet > 9993b2ff88924f90b8990e61f4160ea8 after 1 attempt(s): Failed to write to > server: a7bdcd52b0ad478a8b72ba891ee3ad29 (kudu-185:7050): Write > RPC to 10.110.18.185:7050 timed out after 9.987s (SENT) > > Timed out: Failed to write batch of 16 ops to tablet > f31f017e4b134793b5dc4a9f47e3e45c after 6 attempt(s): Failed to write to > server: (no server available): Write(tablet: f31f017e4b134793b5dc4a9f47e3e45c, > num_ops: 16, num_attempts: 6) passed its deadline: Illegal state: Replica > 80dd197733dd4b80891b4f87a2c12bd6 is not leader of this config. Role: > FOLLOWER. Consensus state: current_term: 9 leader_uuid: "" committed_config > { opid_index: -1 OBSOLETE_local: false peers { permanent_uuid: " > d23fb4ab2e6b478c9849ceea4bf01ccc" member_type: VOTER last_known_addr { > host: "kudu-102" port: 7050 } } peers { permanent_uuid: " > 6f9b8c5dcbf04dc2bfe6b48b419ed46e" member_type: VOTER last_known_addr { > host: "kudu-184" port: 7050 } } peers { permanent_uuid: " > 80dd197733dd4b80891b4f87a2c12bd6" member_type: VOTER last_known_addr { > host: "kudu-185" port: 7050 } } } > > I try to increase --flush_threshold_mb > <http://kudu.apache.org/docs/configuration_reference.html#kudu-master_flush_threshold_mb> > flag, but i get the same erro > how can i avoid the issue ? > > Thanks > > >
