[
https://issues.apache.org/jira/browse/HBASE-16708?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=15668536#comment-15668536
]
Yi Liang edited comment on HBASE-16708 at 11/15/16 10:07 PM:
-------------------------------------------------------------
Hi [~ndimiduk]
Carry your comments to show the full class name, since the full class name
is only stored in a map called coprocessorServiceHandlers in each region,
register each coprocessor, and it is a map store <ServiceName, Real
ServiceClass>. Use my case as example, the coprocessorServiceHandlers will
store <SumService, org.myname.hbase.coprocessor.endpoint.SumEndPoint>, the
SumService is created by xx.proto, and SumEndPoint is the coprocessor class
extend SumService. So the only way to get the full class name is through
coprocessorServiceHandlers. The detail about how to get access to
coprocessorServiceHandlers is in the patch, could you help me to review it.
Thanks
The current output looks like:
{code}
2016-11-15 13:34:22,191 WARN
[RpcServer.FifoWFPBQ.default.handler=59,queue=5,port=16020] ipc.RpcServer:
(responseTooLarge):
{"call":"ExecService(org.apache.hadoop.hbase.shaded.protobuf.generated.ClientProtos$CoprocessorServiceRequest)","starttimems":1479245662059,"responsesize":127,"method":"ExecService","param":"endpoint
coprocessor= class
org.myname.hbase.coprocessor.endpoint.SumEndPoint:getSum","processingtimems":131,"client":"172.16.156.175:44892","queuetimems":1,"class":"HRegionServer"}
{code}
was (Author: easyliangjob):
Hi [~ndimiduk]
Carry your comments to show the full class name, since the full class name
is only stored in a map called coprocessorServiceHandlers in region, register
each coprocessor, and it is a map store <ServiceName, Real ServiceClass>. Use
my case as example, the coprocessorServiceHandlers will store <SumService,
org.myname.hbase.coprocessor.endpoint.SumEndPoint>, the SumService is created
by xx.proto, and SumEndPoint is the coprocessor class extend SumService. So the
only way to get the full class name is through coprocessorServiceHandlers.
The detail about how to get access to coprocessorServiceHandlers is in the
patch, could you help me to review it. Thanks
The current output looks like:
{code}
2016-11-15 13:34:22,191 WARN
[RpcServer.FifoWFPBQ.default.handler=59,queue=5,port=16020] ipc.RpcServer:
(responseTooLarge):
{"call":"ExecService(org.apache.hadoop.hbase.shaded.protobuf.generated.ClientProtos$CoprocessorServiceRequest)","starttimems":1479245662059,"responsesize":127,"method":"ExecService","param":"endpoint
coprocessor= class
org.myname.hbase.coprocessor.endpoint.SumEndPoint:getSum","processingtimems":131,"client":"172.16.156.175:44892","queuetimems":1,"class":"HRegionServer"}
{code}
> Expose endpoint Coprocessor name in "responseTooSlow" log messages
> ------------------------------------------------------------------
>
> Key: HBASE-16708
> URL: https://issues.apache.org/jira/browse/HBASE-16708
> Project: HBase
> Issue Type: Improvement
> Affects Versions: 1.1.2
> Reporter: Nick Dimiduk
> Assignee: Yi Liang
>
> Operational diagnostics of a Phoenix install would be easier if we included
> which endpoint coprocessor was being called in this responseTooSlow WARN
> message.
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
