[ 
https://issues.apache.org/jira/browse/KUDU-3420?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=17633546#comment-17633546
 ] 

ASF subversion and git services commented on KUDU-3420:
-------------------------------------------------------

Commit f88c33ccab88f2e6f9bf8c2155f7115e79645882 in kudu's branch 
refs/heads/master from xinghuayu007
[ https://gitbox.apache.org/repos/asf?p=kudu.git;h=f88c33cca ]

[KUDU-3420] Fixbug of list local replica

Execute command: kudu local_replica list -fs_data_dirs=....
-fs_wal_dir=... --list_detail=true will core. Because
this command inits fs manager without opening block manager,
but it still uses fs mananger to get data dir group PB.

The core stack:
@        0x14ee526 (unknown)
@   0x7f677076f630 (unknown)
@        0x13ba311 kudu::tablet::TabletMetadata::LoadFromSuperBlock()
@        0x13bafa0 kudu::tablet::TabletMetadata::LoadFromDisk()
@        0x13bb2a5 kudu::tablet::TabletMetadata::Load()
@         0xe00d88 kudu::tools::DumpTabletMeta()
@         0xe01417 kudu::tools::ListLocalReplicas()
@         0xdc2824 std::_Function_handler<>::_M_invoke()
@        0x1542350 kudu::tools::Action::Run()
@         0xe774e5 kudu::tools::DispatchCommand()
@         0xe77fb7 kudu::tools::RunTool()
@         0xd2ca13 main
@   0x7f676e960555 __libc_start_main
@         0xdad58

See the detail #KUDU-3420

Change-Id: I7d8551280ad8edd416a0dd1b9611185cabe1a5bc
Reviewed-on: http://gerrit.cloudera.org:8080/19209
Reviewed-by: Yingchun Lai <[email protected]>
Tested-by: Yingchun Lai <[email protected]>


> list local replica with parameter --list_detail=true will core
> --------------------------------------------------------------
>
>                 Key: KUDU-3420
>                 URL: https://issues.apache.org/jira/browse/KUDU-3420
>             Project: Kudu
>          Issue Type: Bug
>            Reporter: Xixu Wang
>            Priority: Major
>         Attachments: image-2022-11-05-23-00-02-955.png
>
>
> Execute command: kudu local_replica list -fs_data_dirs=.... -fs_wal_dir=...
> --list_detail=true will core. Because this command inits fs manager without
> opening block manager, but it uses fs mananger to get data dir group PB.
> Steps to repeat the bug:
> [DEV (v.v) sa_cluster@hybrid01 ~]$ sudo ./kudu  local_replica  list 
> -fs_data_dirs=//kudu/tserver_data -fs_wal_dir=/kudu/tserver_wal --list_detail
> I20221105 21:08:04.123214  9633 fs_manager.cc:334] Metadata directory not 
> provided
> I20221105 21:08:04.123286  9633 fs_manager.cc:340] Using write-ahead log 
> directory (fs_wal_dir) as metadata directory
> W20221105 21:08:04.123649  9633 dir_manager.cc:457] IO error: Could not lock 
> instance file. Make sure that Kudu is not already running and you are not 
> trying to run Kudu with a different user than before: lock 
> /kudu/tserver_data/data/block_manager_instance: Resource temporarily 
> unavailable (error 11)
> W20221105 21:08:04.123663  9633 dir_manager.cc:458] Proceeding without lock
> I20221105 21:08:04.123893  9633 fs_manager.cc:522] Time spent opening 
> directory manager: real 0.000s    user 0.000s    sys 0.000s
> I20221105 21:08:04.123909  9633 fs_manager.cc:576] Opened local filesystem: 
> /kudu/tserver_data,/kudu/tserver_wal
> uuid: "2f2514aa3b534a439b09811cfe712312"
> format_stamp: "Formatted at 2022-11-02 08:59:08 on hybrid01"
> Tablet: 89d78f032c9b4653a6fe6467ae77a450
> *** Aborted at 1667653684 (unix time) try "date -d @1667653684" if you are 
> using GNU date ***
> PC: @                0x0 (unknown)
> *** SIGSEGV (@0x0) received by PID 9633 (TID 0x7f6770fc6a00) from PID 0; 
> stack trace: ***
>     @          0x14ee526 (unknown)
>     @     0x7f677076f630 (unknown)
>     @          0x13ba311 kudu::tablet::TabletMetadata::LoadFromSuperBlock()
>     @          0x13bafa0 kudu::tablet::TabletMetadata::LoadFromDisk()
>     @          0x13bb2a5 kudu::tablet::TabletMetadata::Load()
>     @           0xe00d88 kudu::tools::DumpTabletMeta()
>     @           0xe01417 kudu::tools::ListLocalReplicas()
>     @           0xdc2824 std::_Function_handler<>::_M_invoke()
>     @          0x1542350 kudu::tools::Action::Run()
>     @           0xe774e5 kudu::tools::DispatchCommand()
>     @           0xe77fb7 kudu::tools::RunTool()
>     @           0xd2ca13 main
>     @     0x7f676e960555 __libc_start_main
>     @           0xdad584 (unknown)
> Segmentation fault
>  
> See the core stack:
> !image-2022-11-05-23-00-02-955.png!



--
This message was sent by Atlassian Jira
(v8.20.10#820010)

Reply via email to