[
https://issues.apache.org/jira/browse/HAWQ-1640?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel
]
Radar Lei updated HAWQ-1640:
----------------------------
Affects Version/s: 2.3.0.0-incubating
Fix Version/s: (was: 2.3.0.0-incubating)
backlog
> process not exit after query finished immediately while client connection lost
> ------------------------------------------------------------------------------
>
> Key: HAWQ-1640
> URL: https://issues.apache.org/jira/browse/HAWQ-1640
> Project: Apache HAWQ
> Issue Type: Bug
> Components: Core
> Affects Versions: 2.3.0.0-incubating
> Reporter: TaoJIn
> Assignee: Radar Lei
> Priority: Major
> Fix For: backlog
>
> Attachments: gdb backtrace.jpeg
>
>
> When client (such as pgbouncer,jdbc,zeppelin) connected to hawq and
> execute a long query,if the client connection interrupted before query
> finished,the server process will not exit until an hour later.
> This issue was happend in HAWQ 2.3.0.0-incubating.And
> set parameter gp_interconnect_transmit_timeout to 600(default 3600) will
> reduce the time to 10 minutes.
> When the query wa running,we could see its status in
> pg_stat_activty,but after it finished we could only saw the process id
> in pg_locks and OS process.
> We could saw some error log as below:
> $ tailf hawq-2018-07-04_063514.csv|grep p294
> 2018-07-04 08:13:29.595365
> UTC,"dev","hdb",p294,th1628359104,"172.17.10.148","63974",2018-07-04
> 06:37:28 UTC,58896,con19,cmd32,seg-1,,,x58896,sx1,"LOG","00000","ConnID
> 5. Returned resource to resource manager.",,,,,,,0,,"rmcomm_QD2RM.c",951,
> 2018-07-04 08:13:29.595555
> UTC,"dev","hdb",p294,th1628359104,"172.17.10.148","63974",2018-07-04
> 06:37:28 UTC,58896,con19,cmd32,seg-1,,,x58896,sx1,"LOG","00000","ConnID
> 5. Unregistered from HAWQ resource manager.",,,,,,,0,,"rmcomm_QD2RM.c",661,
> 2018-07-04 08:15:58.706458
> UTC,"dev","hdb",p294,th1628359104,"172.17.10.148","63974",2018-07-04
> 06:37:28 UTC,58903,con19,cmd34,seg-1,,,x58903,sx1,"LOG","00000","ConnID
> 6. Registered in HAWQ resource manager (By OID)",,,,,,"select * from
> cppayorderproduct",0,,"rmcomm_QD2RM.c",609,
> 2018-07-04 08:15:58.706640
> UTC,"dev","hdb",p294,th1628359104,"172.17.10.148","63974",2018-07-04
> 06:37:28 UTC,58903,con19,cmd34,seg-1,,,x58903,sx1,"LOG","00000","ConnID
> 6. Acquired resource from resource manager, (256 MB, 0.062500 CORE) x
> 18.",,,,,,"select * from cppayorderproduct",0,,"rmcomm_QD2RM.c",868,
> 2018-07-04 09:04:56.190873
> UTC,"dev","hdb",p294,th1628359104,"172.17.10.148","63974",2018-07-04
> 06:37:28 UTC,58903,con19,cmd35,seg-1,,,x58903,sx1,"LOG","08006","could
> not send data to client: Connection reset by peer",,,,,,"select * from
> cppayorderproduct",0,,"pqcomm.c",1413,
> 2018-07-04 09:04:56.192347
> UTC,"dev","hdb",p294,th1628359104,"172.17.10.148","63974",2018-07-04
> 06:37:28
> UTC,58903,con19,cmd35,seg-1,,,x58903,sx1,"FATAL","08006","connection to
> client lost",,,,,,"select * from cppayorderproduct",0,,"postgres.c",3606,
> 2018-07-04 10:04:56.306412
> UTC,"dev","hdb",p294,th1627412224,"172.17.10.148","63974",2018-07-04
> 06:37:28
> UTC,58903,con19,cmd35,seg-1,,,x58903,sx1,"LOG","00000","function
> executormgr_consume meets error, connection is bad.",,,,,,,0,,,,
> 2018-07-04 10:04:56.306535
> UTC,"dev","hdb",p294,th1627412224,"172.17.10.148","63974",2018-07-04
> 06:37:28
> UTC,58903,con19,cmd35,seg-1,,,x58903,sx1,"LOG","00000","dispmgt_thread_func_run():
>
> fail to consume data. Will exit and clean up.",,,,,,,0,,,,
> 2018-07-04 10:04:56.309663
> UTC,"dev","hdb",p294,th1627412224,"172.17.10.148","63974",2018-07-04
> 06:37:28
> UTC,58903,con19,cmd35,seg-1,,,x58903,sx1,"LOG","00000","function
> executormgr_cancel calling executormgr_catch_error",,,,,,,0,,,,
> 2018-07-04 10:04:56.312741
> UTC,"dev","hdb",p294,th1627412224,"172.17.10.148","63974",2018-07-04
> 06:37:28
> UTC,58903,con19,cmd35,seg-1,,,x58903,sx1,"LOG","00000","function
> executormgr_cancel calling executormgr_catch_error",,,,,,,0,,,,
> 2018-07-04 10:04:56.315364
> UTC,"dev","hdb",p294,th1627412224,"172.17.10.148","63974",2018-07-04
> 06:37:28
> UTC,58903,con19,cmd35,seg-1,,,x58903,sx1,"LOG","00000","function
> executormgr_cancel calling executormgr_catch_error",,,,,,,0,,,,
> 2018-07-04 10:04:56.317885
> UTC,"dev","hdb",p294,th1627412224,"172.17.10.148","63974",2018-07-04
> 06:37:28
> UTC,58903,con19,cmd35,seg-1,,,x58903,sx1,"LOG","00000","function
> executormgr_cancel calling executormgr_catch_error",,,,,,,0,,,,
> 2018-07-04 10:04:56.320411
> UTC,"dev","hdb",p294,th1627412224,"172.17.10.148","63974",2018-07-04
> 06:37:28
> UTC,58903,con19,cmd35,seg-1,,,x58903,sx1,"LOG","00000","function
> executormgr_cancel calling executormgr_catch_error",,,,,,,0,,,,
> 2018-07-04 10:04:56.322998
> UTC,"dev","hdb",p294,th1627412224,"172.17.10.148","63974",2018-07-04
> 06:37:28
> UTC,58903,con19,cmd35,seg-1,,,x58903,sx1,"LOG","00000","function
> executormgr_cancel calling executormgr_catch_error",,,,,,,0,,,,
> 2018-07-04 10:04:56.327342
> UTC,"dev","hdb",p294,th1627412224,"172.17.10.148","63974",2018-07-04
> 06:37:28
> UTC,58903,con19,cmd35,seg-1,,,x58903,sx1,"LOG","00000","function
> executormgr_cancel calling executormgr_catch_error",,,,,,,0,,,,
> 2018-07-04 10:04:56.330034
> UTC,"dev","hdb",p294,th1627412224,"172.17.10.148","63974",2018-07-04
> 06:37:28
> UTC,58903,con19,cmd35,seg-1,,,x58903,sx1,"LOG","00000","function
> executormgr_cancel calling executormgr_catch_error",,,,,,,0,,,,
> 2018-07-04 10:04:56.332656
> UTC,"dev","hdb",p294,th1627412224,"172.17.10.148","63974",2018-07-04
> 06:37:28
> UTC,58903,con19,cmd35,seg-1,,,x58903,sx1,"LOG","00000","function
> executormgr_cancel calling executormgr_catch_error",,,,,,,0,,,,
> 2018-07-04 10:04:56.335257
> UTC,"dev","hdb",p294,th1627412224,"172.17.10.148","63974",2018-07-04
> 06:37:28
> UTC,58903,con19,cmd35,seg-1,,,x58903,sx1,"LOG","00000","function
> executormgr_cancel calling executormgr_catch_error",,,,,,,0,,,,
> 2018-07-04 10:04:56.337972
> UTC,"dev","hdb",p294,th1627412224,"172.17.10.148","63974",2018-07-04
> 06:37:28
> UTC,58903,con19,cmd35,seg-1,,,x58903,sx1,"LOG","00000","function
> executormgr_cancel calling executormgr_catch_error",,,,,,,0,,,,
> 2018-07-04 10:04:56.340634
> UTC,"dev","hdb",p294,th1627412224,"172.17.10.148","63974",2018-07-04
> 06:37:28
> UTC,58903,con19,cmd35,seg-1,,,x58903,sx1,"LOG","00000","function
> executormgr_cancel calling executormgr_catch_error",,,,,,,0,,,,
> 2018-07-04 10:04:56.343785
> UTC,"dev","hdb",p294,th1627412224,"172.17.10.148","63974",2018-07-04
> 06:37:28
> UTC,58903,con19,cmd35,seg-1,,,x58903,sx1,"LOG","00000","function
> executormgr_cancel calling executormgr_catch_error",,,,,,,0,,,,
> 2018-07-04 10:04:56.346309
> UTC,"dev","hdb",p294,th1627412224,"172.17.10.148","63974",2018-07-04
> 06:37:28
> UTC,58903,con19,cmd35,seg-1,,,x58903,sx1,"LOG","00000","function
> executormgr_cancel calling executormgr_catch_error",,,,,,,0,,,,
> 2018-07-04 10:04:56.350458
> UTC,"dev","hdb",p294,th1627412224,"172.17.10.148","63974",2018-07-04
> 06:37:28
> UTC,58903,con19,cmd35,seg-1,,,x58903,sx1,"LOG","00000","function
> executormgr_cancel calling executormgr_catch_error",,,,,,,0,,,,
> 2018-07-04 10:04:56.352760
> UTC,"dev","hdb",p294,th1627412224,"172.17.10.148","63974",2018-07-04
> 06:37:28
> UTC,58903,con19,cmd35,seg-1,,,x58903,sx1,"LOG","00000","function
> executormgr_cancel calling executormgr_catch_error",,,,,,,0,,,,
> 2018-07-04 10:04:56.354846
> UTC,"dev","hdb",p294,th1627412224,"172.17.10.148","63974",2018-07-04
> 06:37:28
> UTC,58903,con19,cmd35,seg-1,,,x58903,sx1,"LOG","00000","function
> executormgr_cancel calling executormgr_catch_error",,,,,,,0,,,,
> 2018-07-04 10:04:56.354996
> UTC,"dev","hdb",p294,th1628359104,"172.17.10.148","63974",2018-07-04
> 06:37:28
> UTC,58903,con19,cmd35,seg-1,,,x58903,sx1,"LOG","00000","dispatcher
> thinks centos7-datanode3.centos7_hawq_network is
> down.",,,,,,,0,,"rmcomm_QD2RM.c",1209,
> 2018-07-04 10:04:56.355020
> UTC,"dev","hdb",p294,th1628359104,"172.17.10.148","63974",2018-07-04
> 06:37:28
> UTC,58903,con19,cmd35,seg-1,,,x58903,sx1,"LOG","00000","dispatcher sends
> 1 failed host(s) to resource manager.",,,,,,,0,,"rmcomm_QD2RM.c",1213,
> 2018-07-04 10:04:56.355398
> UTC,"dev","hdb",p294,th1628359104,"172.17.10.148","63974",2018-07-04
> 06:37:28 UTC,58903,con19,cmd35,seg-1,,,x58903,sx1,"LOG","00000","succeed
> in sending failed host to resource manager.",,,,,,,0,,"rmcomm_QD2RM.c",1232,
> 2018-07-04 10:04:56.356559
> UTC,"dev","hdb",p294,th1628359104,"172.17.10.148","63974",2018-07-04
> 06:37:28 UTC,58903,con19,cmd35,seg-1,,,x58903,sx1,"LOG","00000","ConnID
> 6. Returned resource to resource manager.",,,,,,,0,,"rmcomm_QD2RM.c",951,
> 2018-07-04 10:04:56.356869
> UTC,"dev","hdb",p294,th1628359104,"172.17.10.148","63974",2018-07-04
> 06:37:28 UTC,58903,con19,cmd35,seg-1,,,x58903,sx1,"LOG","00000","ConnID
> 6. Unregistered from HAWQ resource manager.",,,,,,,0,,"rmcomm_QD2RM.c",661,
> 2018-07-04 10:04:56.356892
> UTC,"dev","hdb",p294,th1628359104,"172.17.10.148","63974",2018-07-04
> 06:37:28 UTC,0,con19,cmd35,seg-1,,,,,"LOG","00000","clean up
> communication to resource manager now.",,,,,,,0,,"rmcomm_QD2RM.c",460,
> 2018-07-04 10:04:56.379198
> UTC,"dev","hdb",p294,th998237952,"172.17.10.148","63974",2018-07-04
> 06:37:28
> UTC,0,con19,cmd35,seg-1,,,,,"LOG","00000","generateResourceRefreshHeartBeat
> exits.",,,,,,,0,,,,
> attched is the server process backtrace
--
This message was sent by Atlassian JIRA
(v7.6.3#76005)
