hbase-issues mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From "Victor Xu (JIRA)" <j...@apache.org>
Subject [jira] [Updated] (HBASE-11654) WAL Splitting dirs are not deleted after replay.
Date Sat, 02 Aug 2014 11:43:11 GMT

     [ https://issues.apache.org/jira/browse/HBASE-11654?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel
]

Victor Xu updated HBASE-11654:
------------------------------

    Attachment: HBASE-11654.patch

After debugging, I found that the master.MasterFileSystem misused the constructor of SplitLogManager.
It used the constructor for tests, which omitted  TaskFinisher, rather than the normal one.
So I made a small patch, and tested on my cluster. It works fine.

> WAL Splitting dirs are not deleted after replay.
> ------------------------------------------------
>
>                 Key: HBASE-11654
>                 URL: https://issues.apache.org/jira/browse/HBASE-11654
>             Project: HBase
>          Issue Type: Bug
>          Components: master, wal
>    Affects Versions: 0.98.4
>            Reporter: Victor Xu
>         Attachments: HBASE-11654.patch
>
>
> I build a small cluster (20 nodes, several hundred regions) with hbase-0.98.4. And I
found some splitting directories in /hbase/WALs/ today, which is very strange because those
logs should have been replayed. Even though the ZK nodes of the dead RS had been deleted,
these splitting directories still can cause a serious trouble for cluster restart. It resplitted
and replayed all the splitting directories every time I restart my cluster, and cost a huge
amount of time. Can't imagine what could happened if it's a cluster with hundreds of nodes
and tens of thousands of regions.
> Found 56 items
> drwxr-xr-x   - hadoop hadoop          0 2014-07-30 18:21 /hbase/WALs/hdpdev1.cm6.tbsite.net,60020,1406714828440-splitting
> drwxr-xr-x   - hadoop hadoop          0 2014-07-30 18:43 /hbase/WALs/hdpdev1.cm6.tbsite.net,60020,1406716991836-splitting
> drwxr-xr-x   - hadoop hadoop          0 2014-08-02 19:16 /hbase/WALs/hdpdev1.cm6.tbsite.net,60020,1406778815585
> drwxr-xr-x   - hadoop hadoop          0 2014-07-30 18:26 /hbase/WALs/hdpdev10.cm6.tbsite.net,60020,1406526862752-splitting
> drwxr-xr-x   - hadoop hadoop          0 2014-07-30 18:43 /hbase/WALs/hdpdev10.cm6.tbsite.net,60020,1406716933471-splitting
> drwxr-xr-x   - hadoop hadoop          0 2014-08-02 19:02 /hbase/WALs/hdpdev10.cm6.tbsite.net,60020,1406778815536
> drwxr-xr-x   - hadoop hadoop          0 2014-07-30 18:26 /hbase/WALs/hdpdev11.cm6.tbsite.net,60020,1406526862802-splitting
> drwxr-xr-x   - hadoop hadoop          0 2014-07-30 18:43 /hbase/WALs/hdpdev11.cm6.tbsite.net,60020,1406716992986-splitting
> drwxr-xr-x   - hadoop hadoop          0 2014-08-02 19:14 /hbase/WALs/hdpdev11.cm6.tbsite.net,60020,1406778815552
> drwxr-xr-x   - hadoop hadoop          0 2014-07-30 18:15 /hbase/WALs/hdpdev12.cm6.tbsite.net,60020,1406526862752-splitting
> drwxr-xr-x   - hadoop hadoop          0 2014-07-30 18:43 /hbase/WALs/hdpdev12.cm6.tbsite.net,60020,1406716992874-splitting
> drwxr-xr-x   - hadoop hadoop          0 2014-08-02 19:20 /hbase/WALs/hdpdev12.cm6.tbsite.net,60020,1406778816074
> drwxr-xr-x   - hadoop hadoop          0 2014-07-30 18:27 /hbase/WALs/hdpdev13.cm6.tbsite.net,60020,1406526862832-splitting
> drwxr-xr-x   - hadoop hadoop          0 2014-07-30 18:43 /hbase/WALs/hdpdev13.cm6.tbsite.net,60020,1406716992753-splitting
> drwxr-xr-x   - hadoop hadoop          0 2014-08-02 19:15 /hbase/WALs/hdpdev13.cm6.tbsite.net,60020,1406857929773
> drwxr-xr-x   - hadoop hadoop          0 2014-07-30 18:26 /hbase/WALs/hdpdev14.cm6.tbsite.net,60020,1406526862736-splitting
> drwxr-xr-x   - hadoop hadoop          0 2014-07-30 18:43 /hbase/WALs/hdpdev14.cm6.tbsite.net,60020,1406716992923-splitting
> drwxr-xr-x   - hadoop hadoop          0 2014-08-02 19:25 /hbase/WALs/hdpdev14.cm6.tbsite.net,60020,1406778815595
> drwxr-xr-x   - hadoop hadoop          0 2014-07-30 18:19 /hbase/WALs/hdpdev15.cm6.tbsite.net,60020,1406526862821-splitting
> drwxr-xr-x   - hadoop hadoop          0 2014-07-30 18:43 /hbase/WALs/hdpdev15.cm6.tbsite.net,60020,1406716993082-splitting
> drwxr-xr-x   - hadoop hadoop          0 2014-08-02 19:15 /hbase/WALs/hdpdev15.cm6.tbsite.net,60020,1406778815578
> drwxr-xr-x   - hadoop hadoop          0 2014-07-30 18:21 /hbase/WALs/hdpdev16.cm6.tbsite.net,60020,1406526862816-splitting
> drwxr-xr-x   - hadoop hadoop          0 2014-07-30 18:43 /hbase/WALs/hdpdev16.cm6.tbsite.net,60020,1406716992787-splitting
> drwxr-xr-x   - hadoop hadoop          0 2014-08-02 19:15 /hbase/WALs/hdpdev16.cm6.tbsite.net,60020,1406778816006
> drwxr-xr-x   - hadoop hadoop          0 2014-07-30 18:43 /hbase/WALs/hdpdev17.cm6.tbsite.net,60020,1406716992814-splitting
> drwxr-xr-x   - hadoop hadoop          0 2014-08-02 19:26 /hbase/WALs/hdpdev17.cm6.tbsite.net,60020,1406778815579
> drwxr-xr-x   - hadoop hadoop          0 2014-07-30 18:43 /hbase/WALs/hdpdev18.cm6.tbsite.net,60020,1406716993051-splitting
> drwxr-xr-x   - hadoop hadoop          0 2014-08-02 19:24 /hbase/WALs/hdpdev18.cm6.tbsite.net,60020,1406778815587
> drwxr-xr-x   - hadoop hadoop          0 2014-07-30 18:23 /hbase/WALs/hdpdev19.cm6.tbsite.net,60020,1406526862720-splitting
> drwxr-xr-x   - hadoop hadoop          0 2014-07-30 18:43 /hbase/WALs/hdpdev19.cm6.tbsite.net,60020,1406716992736-splitting
> drwxr-xr-x   - hadoop hadoop          0 2014-08-02 19:22 /hbase/WALs/hdpdev19.cm6.tbsite.net,60020,1406865567732
> drwxr-xr-x   - hadoop hadoop          0 2014-08-02 19:21 /hbase/WALs/hdpdev2.cm6.tbsite.net,60020,1406778815846
> drwxr-xr-x   - hadoop hadoop          0 2014-07-30 18:27 /hbase/WALs/hdpdev20.cm6.tbsite.net,60020,1406714346484-splitting
> drwxr-xr-x   - hadoop hadoop          0 2014-07-30 18:43 /hbase/WALs/hdpdev20.cm6.tbsite.net,60020,1406716991741-splitting
> drwxr-xr-x   - hadoop hadoop          0 2014-08-02 19:27 /hbase/WALs/hdpdev20.cm6.tbsite.net,60020,1406778815555
> drwxr-xr-x   - hadoop hadoop          0 2014-07-30 18:25 /hbase/WALs/hdpdev3.cm6.tbsite.net,60020,1406714830504-splitting
> drwxr-xr-x   - hadoop hadoop          0 2014-07-30 18:43 /hbase/WALs/hdpdev3.cm6.tbsite.net,60020,1406716992137-splitting
> drwxr-xr-x   - hadoop hadoop          0 2014-08-02 19:24 /hbase/WALs/hdpdev3.cm6.tbsite.net,60020,1406778815585
> drwxr-xr-x   - hadoop hadoop          0 2014-07-30 18:21 /hbase/WALs/hdpdev4.cm6.tbsite.net,60020,1406714829881-splitting
> drwxr-xr-x   - hadoop hadoop          0 2014-07-30 18:43 /hbase/WALs/hdpdev4.cm6.tbsite.net,60020,1406716992118-splitting
> drwxr-xr-x   - hadoop hadoop          0 2014-08-02 19:13 /hbase/WALs/hdpdev4.cm6.tbsite.net,60020,1406864942962
> drwxr-xr-x   - hadoop hadoop          0 2014-07-30 18:10 /hbase/WALs/hdpdev5.cm6.tbsite.net,60020,1406526862790-splitting
> drwxr-xr-x   - hadoop hadoop          0 2014-07-30 18:22 /hbase/WALs/hdpdev5.cm6.tbsite.net,60020,1406715762598-splitting
> drwxr-xr-x   - hadoop hadoop          0 2014-07-30 18:43 /hbase/WALs/hdpdev5.cm6.tbsite.net,60020,1406716991309-splitting
> drwxr-xr-x   - hadoop hadoop          0 2014-07-31 15:39 /hbase/WALs/hdpdev5.cm6.tbsite.net,60020,1406778815529-splitting
> drwxr-xr-x   - hadoop hadoop          0 2014-08-02 19:13 /hbase/WALs/hdpdev5.cm6.tbsite.net,60020,1406941782379
> drwxr-xr-x   - hadoop hadoop          0 2014-07-30 18:43 /hbase/WALs/hdpdev6.cm6.tbsite.net,60020,1406716992903-splitting
> drwxr-xr-x   - hadoop hadoop          0 2014-08-02 19:19 /hbase/WALs/hdpdev6.cm6.tbsite.net,60020,1406778815530
> drwxr-xr-x   - hadoop hadoop          0 2014-07-30 18:24 /hbase/WALs/hdpdev7.cm6.tbsite.net,60020,1406526862796-splitting
> drwxr-xr-x   - hadoop hadoop          0 2014-07-30 18:43 /hbase/WALs/hdpdev7.cm6.tbsite.net,60020,1406716993002-splitting
> drwxr-xr-x   - hadoop hadoop          0 2014-08-02 19:24 /hbase/WALs/hdpdev7.cm6.tbsite.net,60020,1406778815785
> drwxr-xr-x   - hadoop hadoop          0 2014-07-30 18:43 /hbase/WALs/hdpdev8.cm6.tbsite.net,60020,1406716991377-splitting
> drwxr-xr-x   - hadoop hadoop          0 2014-08-02 19:21 /hbase/WALs/hdpdev8.cm6.tbsite.net,60020,1406778815557
> drwxr-xr-x   - hadoop hadoop          0 2014-07-30 18:28 /hbase/WALs/hdpdev9.cm6.tbsite.net,60020,1406716099285-splitting
> drwxr-xr-x   - hadoop hadoop          0 2014-07-30 18:43 /hbase/WALs/hdpdev9.cm6.tbsite.net,60020,1406716991336-splitting
> drwxr-xr-x   - hadoop hadoop          0 2014-08-02 19:10 /hbase/WALs/hdpdev9.cm6.tbsite.net,60020,1406778815554



--
This message was sent by Atlassian JIRA
(v6.2#6252)

Mime
View raw message