You are viewing a plain text version of this content. The canonical link for it is here.
Posted to issues@hbase.apache.org by "Gregory Chanan (JIRA)" <ji...@apache.org> on 2012/06/22 22:31:42 UTC
[jira] [Created] (HBASE-6260) balancer state should be stored in ZK
Gregory Chanan created HBASE-6260:
-------------------------------------
Summary: balancer state should be stored in ZK
Key: HBASE-6260
URL: https://issues.apache.org/jira/browse/HBASE-6260
Project: HBase
Issue Type: Task
Components: master, zookeeper
Affects Versions: 0.96.0
Reporter: Gregory Chanan
Assignee: Gregory Chanan
See: https://issues.apache.org/jira/browse/HBASE-5953?focusedCommentId=13270200&page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel#comment-13270200
And: https://issues.apache.org/jira/browse/HBASE-5630?focusedCommentId=13399225&page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel#comment-13399225
In short, we need to move the balancer state to ZK so that it won't have to be restarted if the master dies.
--
This message is automatically generated by JIRA.
If you think it was sent incorrectly, please contact your JIRA administrators: https://issues.apache.org/jira/secure/ContactAdministrators!default.jspa
For more information on JIRA, see: http://www.atlassian.com/software/jira
[jira] [Updated] (HBASE-6260) balancer state should be stored in ZK
Posted by "Gregory Chanan (JIRA)" <ji...@apache.org>.
[ https://issues.apache.org/jira/browse/HBASE-6260?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ]
Gregory Chanan updated HBASE-6260:
----------------------------------
Priority: Blocker (was: Major)
Marking as blocker for now, if we can figure out how to do it later without breaking clients, it doesn't need to be a blocker.
> balancer state should be stored in ZK
> -------------------------------------
>
> Key: HBASE-6260
> URL: https://issues.apache.org/jira/browse/HBASE-6260
> Project: HBase
> Issue Type: Task
> Components: master, zookeeper
> Affects Versions: 0.96.0
> Reporter: Gregory Chanan
> Assignee: Gregory Chanan
> Priority: Blocker
>
> See: https://issues.apache.org/jira/browse/HBASE-5953?focusedCommentId=13270200&page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel#comment-13270200
> And: https://issues.apache.org/jira/browse/HBASE-5630?focusedCommentId=13399225&page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel#comment-13399225
> In short, we need to move the balancer state to ZK so that it won't have to be restarted if the master dies.
--
This message is automatically generated by JIRA.
If you think it was sent incorrectly, please contact your JIRA administrators
For more information on JIRA, see: http://www.atlassian.com/software/jira
[jira] [Updated] (HBASE-6260) balancer state should be stored in ZK
Posted by "Gregory Chanan (JIRA)" <ji...@apache.org>.
[ https://issues.apache.org/jira/browse/HBASE-6260?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ]
Gregory Chanan updated HBASE-6260:
----------------------------------
Status: Patch Available (was: Reopened)
> balancer state should be stored in ZK
> -------------------------------------
>
> Key: HBASE-6260
> URL: https://issues.apache.org/jira/browse/HBASE-6260
> Project: HBase
> Issue Type: Task
> Components: master, zookeeper
> Affects Versions: 0.96.0
> Reporter: Gregory Chanan
> Assignee: Gregory Chanan
> Priority: Blocker
> Attachments: HBASE-6260-addendum2.patch, HBASE-6260-addendum2.patch, HBASE-6260-addendum.patch, HBASE-6260.patch, HBASE-6260-v2.patch
>
>
> See: https://issues.apache.org/jira/browse/HBASE-5953?focusedCommentId=13270200&page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel#comment-13270200
> And: https://issues.apache.org/jira/browse/HBASE-5630?focusedCommentId=13399225&page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel#comment-13399225
> In short, we need to move the balancer state to ZK so that it won't have to be restarted if the master dies.
--
This message is automatically generated by JIRA.
If you think it was sent incorrectly, please contact your JIRA administrators
For more information on JIRA, see: http://www.atlassian.com/software/jira
[jira] [Commented] (HBASE-6260) balancer state should be stored in
ZK
Posted by "Hudson (JIRA)" <ji...@apache.org>.
[ https://issues.apache.org/jira/browse/HBASE-6260?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13456249#comment-13456249 ]
Hudson commented on HBASE-6260:
-------------------------------
Integrated in HBase-TRUNK-on-Hadoop-2.0.0 #174 (See [https://builds.apache.org/job/HBase-TRUNK-on-Hadoop-2.0.0/174/])
HBASE-6260 ADDENDUM (fix test failure plus comment fixups) (Revision 1384954)
Result = FAILURE
gchanan :
Files :
* /hbase/trunk/hbase-server/src/main/java/org/apache/hadoop/hbase/zookeeper/LoadBalancerTracker.java
> balancer state should be stored in ZK
> -------------------------------------
>
> Key: HBASE-6260
> URL: https://issues.apache.org/jira/browse/HBASE-6260
> Project: HBase
> Issue Type: Task
> Components: master, zookeeper
> Affects Versions: 0.96.0
> Reporter: Gregory Chanan
> Assignee: Gregory Chanan
> Priority: Blocker
> Attachments: 6260-addendum-3.txt, HBASE-6260-addendum2.patch, HBASE-6260-addendum2.patch, HBASE-6260-addendum.patch, HBASE-6260.patch, HBASE-6260-v2.patch
>
>
> See: https://issues.apache.org/jira/browse/HBASE-5953?focusedCommentId=13270200&page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel#comment-13270200
> And: https://issues.apache.org/jira/browse/HBASE-5630?focusedCommentId=13399225&page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel#comment-13399225
> In short, we need to move the balancer state to ZK so that it won't have to be restarted if the master dies.
--
This message is automatically generated by JIRA.
If you think it was sent incorrectly, please contact your JIRA administrators
For more information on JIRA, see: http://www.atlassian.com/software/jira
[jira] [Commented] (HBASE-6260) balancer state should be stored in
ZK
Posted by "Hudson (JIRA)" <ji...@apache.org>.
[ https://issues.apache.org/jira/browse/HBASE-6260?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13456257#comment-13456257 ]
Hudson commented on HBASE-6260:
-------------------------------
Integrated in HBase-TRUNK #3333 (See [https://builds.apache.org/job/HBase-TRUNK/3333/])
HBASE-6260 ADDENDUM (fix test failure plus comment fixups) (Revision 1384954)
Result = FAILURE
gchanan :
Files :
* /hbase/trunk/hbase-server/src/main/java/org/apache/hadoop/hbase/zookeeper/LoadBalancerTracker.java
> balancer state should be stored in ZK
> -------------------------------------
>
> Key: HBASE-6260
> URL: https://issues.apache.org/jira/browse/HBASE-6260
> Project: HBase
> Issue Type: Task
> Components: master, zookeeper
> Affects Versions: 0.96.0
> Reporter: Gregory Chanan
> Assignee: Gregory Chanan
> Priority: Blocker
> Attachments: 6260-addendum-3.txt, HBASE-6260-addendum2.patch, HBASE-6260-addendum2.patch, HBASE-6260-addendum.patch, HBASE-6260.patch, HBASE-6260-v2.patch
>
>
> See: https://issues.apache.org/jira/browse/HBASE-5953?focusedCommentId=13270200&page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel#comment-13270200
> And: https://issues.apache.org/jira/browse/HBASE-5630?focusedCommentId=13399225&page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel#comment-13399225
> In short, we need to move the balancer state to ZK so that it won't have to be restarted if the master dies.
--
This message is automatically generated by JIRA.
If you think it was sent incorrectly, please contact your JIRA administrators
For more information on JIRA, see: http://www.atlassian.com/software/jira
[jira] [Updated] (HBASE-6260) balancer state should be stored in ZK
Posted by "Gregory Chanan (JIRA)" <ji...@apache.org>.
[ https://issues.apache.org/jira/browse/HBASE-6260?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ]
Gregory Chanan updated HBASE-6260:
----------------------------------
Attachment: HBASE-6260.patch
* Attached HBASE-6260.patch *
I haven't run this through the test suite yet, let's see what HadoopQA says.
- Master reads balanceSwitch state from ZK
- Adds two test cases:
1. If balancer is on and master dies, and new master takes over, balancer is still running
2. 1. If balancer is off and master dies, and new master takes over, balancer is still not running
> balancer state should be stored in ZK
> -------------------------------------
>
> Key: HBASE-6260
> URL: https://issues.apache.org/jira/browse/HBASE-6260
> Project: HBase
> Issue Type: Task
> Components: master, zookeeper
> Affects Versions: 0.96.0
> Reporter: Gregory Chanan
> Assignee: Gregory Chanan
> Priority: Blocker
> Attachments: HBASE-6260.patch
>
>
> See: https://issues.apache.org/jira/browse/HBASE-5953?focusedCommentId=13270200&page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel#comment-13270200
> And: https://issues.apache.org/jira/browse/HBASE-5630?focusedCommentId=13399225&page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel#comment-13399225
> In short, we need to move the balancer state to ZK so that it won't have to be restarted if the master dies.
--
This message is automatically generated by JIRA.
If you think it was sent incorrectly, please contact your JIRA administrators
For more information on JIRA, see: http://www.atlassian.com/software/jira
[jira] [Commented] (HBASE-6260) balancer state should be stored in
ZK
Posted by "Hudson (JIRA)" <ji...@apache.org>.
[ https://issues.apache.org/jira/browse/HBASE-6260?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13455510#comment-13455510 ]
Hudson commented on HBASE-6260:
-------------------------------
Integrated in HBase-TRUNK-on-Hadoop-2.0.0 #172 (See [https://builds.apache.org/job/HBase-TRUNK-on-Hadoop-2.0.0/172/])
HBASE-6260 balancer state should be stored in ZK (Revision 1384593)
Result = FAILURE
gchanan :
Files :
* /hbase/trunk/hbase-server/src/main/java/org/apache/hadoop/hbase/master/HMaster.java
* /hbase/trunk/hbase-server/src/main/java/org/apache/hadoop/hbase/protobuf/generated/LoadBalancerProtos.java
* /hbase/trunk/hbase-server/src/main/java/org/apache/hadoop/hbase/zookeeper/LoadBalancerTracker.java
* /hbase/trunk/hbase-server/src/main/java/org/apache/hadoop/hbase/zookeeper/ZooKeeperWatcher.java
* /hbase/trunk/hbase-server/src/main/protobuf/LoadBalancer.proto
* /hbase/trunk/hbase-server/src/test/java/org/apache/hadoop/hbase/master/TestMasterFailover.java
> balancer state should be stored in ZK
> -------------------------------------
>
> Key: HBASE-6260
> URL: https://issues.apache.org/jira/browse/HBASE-6260
> Project: HBase
> Issue Type: Task
> Components: master, zookeeper
> Affects Versions: 0.96.0
> Reporter: Gregory Chanan
> Assignee: Gregory Chanan
> Priority: Blocker
> Attachments: HBASE-6260-addendum.patch, HBASE-6260.patch, HBASE-6260-v2.patch
>
>
> See: https://issues.apache.org/jira/browse/HBASE-5953?focusedCommentId=13270200&page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel#comment-13270200
> And: https://issues.apache.org/jira/browse/HBASE-5630?focusedCommentId=13399225&page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel#comment-13399225
> In short, we need to move the balancer state to ZK so that it won't have to be restarted if the master dies.
--
This message is automatically generated by JIRA.
If you think it was sent incorrectly, please contact your JIRA administrators
For more information on JIRA, see: http://www.atlassian.com/software/jira
[jira] [Updated] (HBASE-6260) balancer state should be stored in ZK
Posted by "Gregory Chanan (JIRA)" <ji...@apache.org>.
[ https://issues.apache.org/jira/browse/HBASE-6260?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ]
Gregory Chanan updated HBASE-6260:
----------------------------------
Status: Patch Available (was: Open)
> balancer state should be stored in ZK
> -------------------------------------
>
> Key: HBASE-6260
> URL: https://issues.apache.org/jira/browse/HBASE-6260
> Project: HBase
> Issue Type: Task
> Components: master, zookeeper
> Affects Versions: 0.96.0
> Reporter: Gregory Chanan
> Assignee: Gregory Chanan
> Priority: Blocker
> Attachments: HBASE-6260.patch
>
>
> See: https://issues.apache.org/jira/browse/HBASE-5953?focusedCommentId=13270200&page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel#comment-13270200
> And: https://issues.apache.org/jira/browse/HBASE-5630?focusedCommentId=13399225&page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel#comment-13399225
> In short, we need to move the balancer state to ZK so that it won't have to be restarted if the master dies.
--
This message is automatically generated by JIRA.
If you think it was sent incorrectly, please contact your JIRA administrators
For more information on JIRA, see: http://www.atlassian.com/software/jira
[jira] [Commented] (HBASE-6260) balancer state should be stored in
ZK
Posted by "Hudson (JIRA)" <ji...@apache.org>.
[ https://issues.apache.org/jira/browse/HBASE-6260?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13455516#comment-13455516 ]
Hudson commented on HBASE-6260:
-------------------------------
Integrated in HBase-TRUNK #3329 (See [https://builds.apache.org/job/HBase-TRUNK/3329/])
HBASE-6260 balancer state should be stored in ZK (Revision 1384593)
Result = FAILURE
gchanan :
Files :
* /hbase/trunk/hbase-server/src/main/java/org/apache/hadoop/hbase/master/HMaster.java
* /hbase/trunk/hbase-server/src/main/java/org/apache/hadoop/hbase/protobuf/generated/LoadBalancerProtos.java
* /hbase/trunk/hbase-server/src/main/java/org/apache/hadoop/hbase/zookeeper/LoadBalancerTracker.java
* /hbase/trunk/hbase-server/src/main/java/org/apache/hadoop/hbase/zookeeper/ZooKeeperWatcher.java
* /hbase/trunk/hbase-server/src/main/protobuf/LoadBalancer.proto
* /hbase/trunk/hbase-server/src/test/java/org/apache/hadoop/hbase/master/TestMasterFailover.java
> balancer state should be stored in ZK
> -------------------------------------
>
> Key: HBASE-6260
> URL: https://issues.apache.org/jira/browse/HBASE-6260
> Project: HBase
> Issue Type: Task
> Components: master, zookeeper
> Affects Versions: 0.96.0
> Reporter: Gregory Chanan
> Assignee: Gregory Chanan
> Priority: Blocker
> Attachments: HBASE-6260-addendum.patch, HBASE-6260.patch, HBASE-6260-v2.patch
>
>
> See: https://issues.apache.org/jira/browse/HBASE-5953?focusedCommentId=13270200&page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel#comment-13270200
> And: https://issues.apache.org/jira/browse/HBASE-5630?focusedCommentId=13399225&page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel#comment-13399225
> In short, we need to move the balancer state to ZK so that it won't have to be restarted if the master dies.
--
This message is automatically generated by JIRA.
If you think it was sent incorrectly, please contact your JIRA administrators
For more information on JIRA, see: http://www.atlassian.com/software/jira
[jira] [Commented] (HBASE-6260) balancer state should be stored in
ZK
Posted by "Gregory Chanan (JIRA)" <ji...@apache.org>.
[ https://issues.apache.org/jira/browse/HBASE-6260?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13455588#comment-13455588 ]
Gregory Chanan commented on HBASE-6260:
---------------------------------------
Attached addendum2 which fixes the test issue.
> balancer state should be stored in ZK
> -------------------------------------
>
> Key: HBASE-6260
> URL: https://issues.apache.org/jira/browse/HBASE-6260
> Project: HBase
> Issue Type: Task
> Components: master, zookeeper
> Affects Versions: 0.96.0
> Reporter: Gregory Chanan
> Assignee: Gregory Chanan
> Priority: Blocker
> Attachments: HBASE-6260-addendum2.patch, HBASE-6260-addendum.patch, HBASE-6260.patch, HBASE-6260-v2.patch
>
>
> See: https://issues.apache.org/jira/browse/HBASE-5953?focusedCommentId=13270200&page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel#comment-13270200
> And: https://issues.apache.org/jira/browse/HBASE-5630?focusedCommentId=13399225&page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel#comment-13399225
> In short, we need to move the balancer state to ZK so that it won't have to be restarted if the master dies.
--
This message is automatically generated by JIRA.
If you think it was sent incorrectly, please contact your JIRA administrators
For more information on JIRA, see: http://www.atlassian.com/software/jira
[jira] [Commented] (HBASE-6260) balancer state should be stored in
ZK
Posted by "Ted Yu (JIRA)" <ji...@apache.org>.
[ https://issues.apache.org/jira/browse/HBASE-6260?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13455530#comment-13455530 ]
Ted Yu commented on HBASE-6260:
-------------------------------
Addendum looks good.
I saw the following in both trunk builds, including HBase-TRUNK-on-Hadoop-2.0.0:
{code}
Failed tests: testRegionTransitionOperations(org.apache.hadoop.hbase.coprocessor.TestMasterObserver): Coprocessor should be called on region rebalancing
{code}
Please fix the above.
> balancer state should be stored in ZK
> -------------------------------------
>
> Key: HBASE-6260
> URL: https://issues.apache.org/jira/browse/HBASE-6260
> Project: HBase
> Issue Type: Task
> Components: master, zookeeper
> Affects Versions: 0.96.0
> Reporter: Gregory Chanan
> Assignee: Gregory Chanan
> Priority: Blocker
> Attachments: HBASE-6260-addendum.patch, HBASE-6260.patch, HBASE-6260-v2.patch
>
>
> See: https://issues.apache.org/jira/browse/HBASE-5953?focusedCommentId=13270200&page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel#comment-13270200
> And: https://issues.apache.org/jira/browse/HBASE-5630?focusedCommentId=13399225&page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel#comment-13399225
> In short, we need to move the balancer state to ZK so that it won't have to be restarted if the master dies.
--
This message is automatically generated by JIRA.
If you think it was sent incorrectly, please contact your JIRA administrators
For more information on JIRA, see: http://www.atlassian.com/software/jira
[jira] [Commented] (HBASE-6260) balancer state should be stored in
ZK
Posted by "stack (JIRA)" <ji...@apache.org>.
[ https://issues.apache.org/jira/browse/HBASE-6260?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13455435#comment-13455435 ]
stack commented on HBASE-6260:
------------------------------
Just commit G.
> balancer state should be stored in ZK
> -------------------------------------
>
> Key: HBASE-6260
> URL: https://issues.apache.org/jira/browse/HBASE-6260
> Project: HBase
> Issue Type: Task
> Components: master, zookeeper
> Affects Versions: 0.96.0
> Reporter: Gregory Chanan
> Assignee: Gregory Chanan
> Priority: Blocker
> Attachments: HBASE-6260-addendum.patch, HBASE-6260.patch, HBASE-6260-v2.patch
>
>
> See: https://issues.apache.org/jira/browse/HBASE-5953?focusedCommentId=13270200&page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel#comment-13270200
> And: https://issues.apache.org/jira/browse/HBASE-5630?focusedCommentId=13399225&page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel#comment-13399225
> In short, we need to move the balancer state to ZK so that it won't have to be restarted if the master dies.
--
This message is automatically generated by JIRA.
If you think it was sent incorrectly, please contact your JIRA administrators
For more information on JIRA, see: http://www.atlassian.com/software/jira
[jira] [Updated] (HBASE-6260) balancer state should be stored in ZK
Posted by "Gregory Chanan (JIRA)" <ji...@apache.org>.
[ https://issues.apache.org/jira/browse/HBASE-6260?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ]
Gregory Chanan updated HBASE-6260:
----------------------------------
Attachment: HBASE-6260-addendum2.patch
Reattaching to try to kick off HadoopQA again.
> balancer state should be stored in ZK
> -------------------------------------
>
> Key: HBASE-6260
> URL: https://issues.apache.org/jira/browse/HBASE-6260
> Project: HBase
> Issue Type: Task
> Components: master, zookeeper
> Affects Versions: 0.96.0
> Reporter: Gregory Chanan
> Assignee: Gregory Chanan
> Priority: Blocker
> Attachments: HBASE-6260-addendum2.patch, HBASE-6260-addendum2.patch, HBASE-6260-addendum.patch, HBASE-6260.patch, HBASE-6260-v2.patch
>
>
> See: https://issues.apache.org/jira/browse/HBASE-5953?focusedCommentId=13270200&page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel#comment-13270200
> And: https://issues.apache.org/jira/browse/HBASE-5630?focusedCommentId=13399225&page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel#comment-13399225
> In short, we need to move the balancer state to ZK so that it won't have to be restarted if the master dies.
--
This message is automatically generated by JIRA.
If you think it was sent incorrectly, please contact your JIRA administrators
For more information on JIRA, see: http://www.atlassian.com/software/jira
[jira] [Updated] (HBASE-6260) balancer state should be stored in ZK
Posted by "Gregory Chanan (JIRA)" <ji...@apache.org>.
[ https://issues.apache.org/jira/browse/HBASE-6260?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ]
Gregory Chanan updated HBASE-6260:
----------------------------------
Resolution: Fixed
Status: Resolved (was: Patch Available)
Thanks for the review, Ted.
Committed addendum-v3 to trunk.
> balancer state should be stored in ZK
> -------------------------------------
>
> Key: HBASE-6260
> URL: https://issues.apache.org/jira/browse/HBASE-6260
> Project: HBase
> Issue Type: Task
> Components: master, zookeeper
> Affects Versions: 0.96.0
> Reporter: Gregory Chanan
> Assignee: Gregory Chanan
> Priority: Blocker
> Attachments: 6260-addendum-3.txt, HBASE-6260-addendum2.patch, HBASE-6260-addendum2.patch, HBASE-6260-addendum.patch, HBASE-6260.patch, HBASE-6260-v2.patch
>
>
> See: https://issues.apache.org/jira/browse/HBASE-5953?focusedCommentId=13270200&page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel#comment-13270200
> And: https://issues.apache.org/jira/browse/HBASE-5630?focusedCommentId=13399225&page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel#comment-13399225
> In short, we need to move the balancer state to ZK so that it won't have to be restarted if the master dies.
--
This message is automatically generated by JIRA.
If you think it was sent incorrectly, please contact your JIRA administrators
For more information on JIRA, see: http://www.atlassian.com/software/jira
[jira] [Updated] (HBASE-6260) balancer state should be stored in ZK
Posted by "Gregory Chanan (JIRA)" <ji...@apache.org>.
[ https://issues.apache.org/jira/browse/HBASE-6260?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ]
Gregory Chanan updated HBASE-6260:
----------------------------------
Attachment: HBASE-6260-v2.patch
* Attached HBASE-6260-v2.patch *
Add PB_MAGIC prefix.
> balancer state should be stored in ZK
> -------------------------------------
>
> Key: HBASE-6260
> URL: https://issues.apache.org/jira/browse/HBASE-6260
> Project: HBase
> Issue Type: Task
> Components: master, zookeeper
> Affects Versions: 0.96.0
> Reporter: Gregory Chanan
> Assignee: Gregory Chanan
> Priority: Blocker
> Attachments: HBASE-6260.patch, HBASE-6260-v2.patch
>
>
> See: https://issues.apache.org/jira/browse/HBASE-5953?focusedCommentId=13270200&page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel#comment-13270200
> And: https://issues.apache.org/jira/browse/HBASE-5630?focusedCommentId=13399225&page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel#comment-13399225
> In short, we need to move the balancer state to ZK so that it won't have to be restarted if the master dies.
--
This message is automatically generated by JIRA.
If you think it was sent incorrectly, please contact your JIRA administrators
For more information on JIRA, see: http://www.atlassian.com/software/jira
[jira] [Commented] (HBASE-6260) balancer state should be stored in
ZK
Posted by "Gregory Chanan (JIRA)" <ji...@apache.org>.
[ https://issues.apache.org/jira/browse/HBASE-6260?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13454577#comment-13454577 ]
Gregory Chanan commented on HBASE-6260:
---------------------------------------
@stack:
- I don't know what's going on with ClientProtos.java. I've noticed that a few times when I regenerate, the generated code changes. I'll leave it out.
- I think PB_MAGIC is a good idea. I noticed without it, an empty (newly-created) znode is valid for the purposes of mergeFrom. Not likely to happen, but probably best to have it throw an error in that case.
> balancer state should be stored in ZK
> -------------------------------------
>
> Key: HBASE-6260
> URL: https://issues.apache.org/jira/browse/HBASE-6260
> Project: HBase
> Issue Type: Task
> Components: master, zookeeper
> Affects Versions: 0.96.0
> Reporter: Gregory Chanan
> Assignee: Gregory Chanan
> Priority: Blocker
> Attachments: HBASE-6260.patch
>
>
> See: https://issues.apache.org/jira/browse/HBASE-5953?focusedCommentId=13270200&page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel#comment-13270200
> And: https://issues.apache.org/jira/browse/HBASE-5630?focusedCommentId=13399225&page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel#comment-13399225
> In short, we need to move the balancer state to ZK so that it won't have to be restarted if the master dies.
--
This message is automatically generated by JIRA.
If you think it was sent incorrectly, please contact your JIRA administrators
For more information on JIRA, see: http://www.atlassian.com/software/jira
[jira] [Commented] (HBASE-6260) balancer state should be stored in
ZK
Posted by "Gregory Chanan (JIRA)" <ji...@apache.org>.
[ https://issues.apache.org/jira/browse/HBASE-6260?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13399688#comment-13399688 ]
Gregory Chanan commented on HBASE-6260:
---------------------------------------
@Jon:
Have you seen HBASE-5953?
> balancer state should be stored in ZK
> -------------------------------------
>
> Key: HBASE-6260
> URL: https://issues.apache.org/jira/browse/HBASE-6260
> Project: HBase
> Issue Type: Task
> Components: master, zookeeper
> Affects Versions: 0.96.0
> Reporter: Gregory Chanan
> Assignee: Gregory Chanan
>
> See: https://issues.apache.org/jira/browse/HBASE-5953?focusedCommentId=13270200&page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel#comment-13270200
> And: https://issues.apache.org/jira/browse/HBASE-5630?focusedCommentId=13399225&page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel#comment-13399225
> In short, we need to move the balancer state to ZK so that it won't have to be restarted if the master dies.
--
This message is automatically generated by JIRA.
If you think it was sent incorrectly, please contact your JIRA administrators: https://issues.apache.org/jira/secure/ContactAdministrators!default.jspa
For more information on JIRA, see: http://www.atlassian.com/software/jira
[jira] [Commented] (HBASE-6260) balancer state should be stored in
ZK
Posted by "Hadoop QA (JIRA)" <ji...@apache.org>.
[ https://issues.apache.org/jira/browse/HBASE-6260?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13455987#comment-13455987 ]
Hadoop QA commented on HBASE-6260:
----------------------------------
-1 overall. Here are the results of testing the latest attachment
http://issues.apache.org/jira/secure/attachment/12545168/6260-addendum-3.txt
against trunk revision .
+1 @author. The patch does not contain any @author tags.
-1 tests included. The patch doesn't appear to include any new or modified tests.
Please justify why no new tests are needed for this patch.
Also please list what manual steps were performed to verify this patch.
+1 hadoop2.0. The patch compiles against the hadoop 2.0 profile.
+1 javadoc. The javadoc tool did not generate any warning messages.
-1 javac. The patch appears to cause mvn compile goal to fail.
-1 findbugs. The patch appears to cause Findbugs (version 1.3.9) to fail.
+1 release audit. The applied patch does not increase the total number of release audit warnings.
-1 core tests. The patch failed these unit tests:
Test results: https://builds.apache.org/job/PreCommit-HBASE-Build/2870//testReport/
Console output: https://builds.apache.org/job/PreCommit-HBASE-Build/2870//console
This message is automatically generated.
> balancer state should be stored in ZK
> -------------------------------------
>
> Key: HBASE-6260
> URL: https://issues.apache.org/jira/browse/HBASE-6260
> Project: HBase
> Issue Type: Task
> Components: master, zookeeper
> Affects Versions: 0.96.0
> Reporter: Gregory Chanan
> Assignee: Gregory Chanan
> Priority: Blocker
> Attachments: 6260-addendum-3.txt, HBASE-6260-addendum2.patch, HBASE-6260-addendum2.patch, HBASE-6260-addendum.patch, HBASE-6260.patch, HBASE-6260-v2.patch
>
>
> See: https://issues.apache.org/jira/browse/HBASE-5953?focusedCommentId=13270200&page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel#comment-13270200
> And: https://issues.apache.org/jira/browse/HBASE-5630?focusedCommentId=13399225&page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel#comment-13399225
> In short, we need to move the balancer state to ZK so that it won't have to be restarted if the master dies.
--
This message is automatically generated by JIRA.
If you think it was sent incorrectly, please contact your JIRA administrators
For more information on JIRA, see: http://www.atlassian.com/software/jira
[jira] [Commented] (HBASE-6260) balancer state should be stored in
ZK
Posted by "Jonathan Hsieh (JIRA)" <ji...@apache.org>.
[ https://issues.apache.org/jira/browse/HBASE-6260?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13399692#comment-13399692 ]
Jonathan Hsieh commented on HBASE-6260:
---------------------------------------
@Greg
Did not. Now that I have, its great that it has been done!
> balancer state should be stored in ZK
> -------------------------------------
>
> Key: HBASE-6260
> URL: https://issues.apache.org/jira/browse/HBASE-6260
> Project: HBase
> Issue Type: Task
> Components: master, zookeeper
> Affects Versions: 0.96.0
> Reporter: Gregory Chanan
> Assignee: Gregory Chanan
>
> See: https://issues.apache.org/jira/browse/HBASE-5953?focusedCommentId=13270200&page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel#comment-13270200
> And: https://issues.apache.org/jira/browse/HBASE-5630?focusedCommentId=13399225&page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel#comment-13399225
> In short, we need to move the balancer state to ZK so that it won't have to be restarted if the master dies.
--
This message is automatically generated by JIRA.
If you think it was sent incorrectly, please contact your JIRA administrators: https://issues.apache.org/jira/secure/ContactAdministrators!default.jspa
For more information on JIRA, see: http://www.atlassian.com/software/jira
[jira] [Commented] (HBASE-6260) balancer state should be stored in
ZK
Posted by "Ted Yu (JIRA)" <ji...@apache.org>.
[ https://issues.apache.org/jira/browse/HBASE-6260?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13456197#comment-13456197 ]
Ted Yu commented on HBASE-6260:
-------------------------------
Thanks for being careful.
I think addendum v3 is ready to go.
> balancer state should be stored in ZK
> -------------------------------------
>
> Key: HBASE-6260
> URL: https://issues.apache.org/jira/browse/HBASE-6260
> Project: HBase
> Issue Type: Task
> Components: master, zookeeper
> Affects Versions: 0.96.0
> Reporter: Gregory Chanan
> Assignee: Gregory Chanan
> Priority: Blocker
> Attachments: 6260-addendum-3.txt, HBASE-6260-addendum2.patch, HBASE-6260-addendum2.patch, HBASE-6260-addendum.patch, HBASE-6260.patch, HBASE-6260-v2.patch
>
>
> See: https://issues.apache.org/jira/browse/HBASE-5953?focusedCommentId=13270200&page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel#comment-13270200
> And: https://issues.apache.org/jira/browse/HBASE-5630?focusedCommentId=13399225&page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel#comment-13399225
> In short, we need to move the balancer state to ZK so that it won't have to be restarted if the master dies.
--
This message is automatically generated by JIRA.
If you think it was sent incorrectly, please contact your JIRA administrators
For more information on JIRA, see: http://www.atlassian.com/software/jira
[jira] [Commented] (HBASE-6260) balancer state should be stored in
ZK
Posted by "Ted Yu (JIRA)" <ji...@apache.org>.
[ https://issues.apache.org/jira/browse/HBASE-6260?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13455414#comment-13455414 ]
Ted Yu commented on HBASE-6260:
-------------------------------
{code}
+ * Tracks the load balancer switch up in ZK
{code}
Probably you should make the above comment more general - LoadBalancerTracker may cover more than switch status.
{code}
+ // is data in ZK is null, use default of on.
{code}
Typo: 'is' -> 'if'
> balancer state should be stored in ZK
> -------------------------------------
>
> Key: HBASE-6260
> URL: https://issues.apache.org/jira/browse/HBASE-6260
> Project: HBase
> Issue Type: Task
> Components: master, zookeeper
> Affects Versions: 0.96.0
> Reporter: Gregory Chanan
> Assignee: Gregory Chanan
> Priority: Blocker
> Attachments: HBASE-6260.patch, HBASE-6260-v2.patch
>
>
> See: https://issues.apache.org/jira/browse/HBASE-5953?focusedCommentId=13270200&page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel#comment-13270200
> And: https://issues.apache.org/jira/browse/HBASE-5630?focusedCommentId=13399225&page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel#comment-13399225
> In short, we need to move the balancer state to ZK so that it won't have to be restarted if the master dies.
--
This message is automatically generated by JIRA.
If you think it was sent incorrectly, please contact your JIRA administrators
For more information on JIRA, see: http://www.atlassian.com/software/jira
[jira] [Updated] (HBASE-6260) balancer state should be stored in ZK
Posted by "Ted Yu (JIRA)" <ji...@apache.org>.
[ https://issues.apache.org/jira/browse/HBASE-6260?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ]
Ted Yu updated HBASE-6260:
--------------------------
Attachment: 6260-addendum-3.txt
Addendum v3 combines the first two addenda
> balancer state should be stored in ZK
> -------------------------------------
>
> Key: HBASE-6260
> URL: https://issues.apache.org/jira/browse/HBASE-6260
> Project: HBase
> Issue Type: Task
> Components: master, zookeeper
> Affects Versions: 0.96.0
> Reporter: Gregory Chanan
> Assignee: Gregory Chanan
> Priority: Blocker
> Attachments: 6260-addendum-3.txt, HBASE-6260-addendum2.patch, HBASE-6260-addendum2.patch, HBASE-6260-addendum.patch, HBASE-6260.patch, HBASE-6260-v2.patch
>
>
> See: https://issues.apache.org/jira/browse/HBASE-5953?focusedCommentId=13270200&page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel#comment-13270200
> And: https://issues.apache.org/jira/browse/HBASE-5630?focusedCommentId=13399225&page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel#comment-13399225
> In short, we need to move the balancer state to ZK so that it won't have to be restarted if the master dies.
--
This message is automatically generated by JIRA.
If you think it was sent incorrectly, please contact your JIRA administrators
For more information on JIRA, see: http://www.atlassian.com/software/jira
[jira] [Commented] (HBASE-6260) balancer state should be stored in
ZK
Posted by "stack (JIRA)" <ji...@apache.org>.
[ https://issues.apache.org/jira/browse/HBASE-6260?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13455394#comment-13455394 ]
stack commented on HBASE-6260:
------------------------------
+1
Checked PB_MAGIC is included serializing and managed deserializing. Looks good.
> balancer state should be stored in ZK
> -------------------------------------
>
> Key: HBASE-6260
> URL: https://issues.apache.org/jira/browse/HBASE-6260
> Project: HBase
> Issue Type: Task
> Components: master, zookeeper
> Affects Versions: 0.96.0
> Reporter: Gregory Chanan
> Assignee: Gregory Chanan
> Priority: Blocker
> Attachments: HBASE-6260.patch, HBASE-6260-v2.patch
>
>
> See: https://issues.apache.org/jira/browse/HBASE-5953?focusedCommentId=13270200&page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel#comment-13270200
> And: https://issues.apache.org/jira/browse/HBASE-5630?focusedCommentId=13399225&page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel#comment-13399225
> In short, we need to move the balancer state to ZK so that it won't have to be restarted if the master dies.
--
This message is automatically generated by JIRA.
If you think it was sent incorrectly, please contact your JIRA administrators
For more information on JIRA, see: http://www.atlassian.com/software/jira
[jira] [Commented] (HBASE-6260) balancer state should be stored in
ZK
Posted by "Gregory Chanan (JIRA)" <ji...@apache.org>.
[ https://issues.apache.org/jira/browse/HBASE-6260?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13455367#comment-13455367 ]
Gregory Chanan commented on HBASE-6260:
---------------------------------------
I ran TestForceCacheImportantBlocks locally and it passed. TestReplication has been failing for a long time.
> balancer state should be stored in ZK
> -------------------------------------
>
> Key: HBASE-6260
> URL: https://issues.apache.org/jira/browse/HBASE-6260
> Project: HBase
> Issue Type: Task
> Components: master, zookeeper
> Affects Versions: 0.96.0
> Reporter: Gregory Chanan
> Assignee: Gregory Chanan
> Priority: Blocker
> Attachments: HBASE-6260.patch, HBASE-6260-v2.patch
>
>
> See: https://issues.apache.org/jira/browse/HBASE-5953?focusedCommentId=13270200&page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel#comment-13270200
> And: https://issues.apache.org/jira/browse/HBASE-5630?focusedCommentId=13399225&page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel#comment-13399225
> In short, we need to move the balancer state to ZK so that it won't have to be restarted if the master dies.
--
This message is automatically generated by JIRA.
If you think it was sent incorrectly, please contact your JIRA administrators
For more information on JIRA, see: http://www.atlassian.com/software/jira
[jira] [Commented] (HBASE-6260) balancer state should be stored in
ZK
Posted by "Hadoop QA (JIRA)" <ji...@apache.org>.
[ https://issues.apache.org/jira/browse/HBASE-6260?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13455644#comment-13455644 ]
Hadoop QA commented on HBASE-6260:
----------------------------------
-1 overall. Here are the results of testing the latest attachment
http://issues.apache.org/jira/secure/attachment/12545107/HBASE-6260-addendum2.patch
against trunk revision .
+1 @author. The patch does not contain any @author tags.
-1 tests included. The patch doesn't appear to include any new or modified tests.
Please justify why no new tests are needed for this patch.
Also please list what manual steps were performed to verify this patch.
+1 hadoop2.0. The patch compiles against the hadoop 2.0 profile.
+1 javadoc. The javadoc tool did not generate any warning messages.
-1 javac. The patch appears to cause mvn compile goal to fail.
-1 findbugs. The patch appears to cause Findbugs (version 1.3.9) to fail.
+1 release audit. The applied patch does not increase the total number of release audit warnings.
-1 core tests. The patch failed these unit tests:
org.apache.hadoop.hbase.regionserver.TestAtomicOperation
org.apache.hadoop.hbase.replication.TestReplication
Test results: https://builds.apache.org/job/PreCommit-HBASE-Build/2867//testReport/
Console output: https://builds.apache.org/job/PreCommit-HBASE-Build/2867//console
This message is automatically generated.
> balancer state should be stored in ZK
> -------------------------------------
>
> Key: HBASE-6260
> URL: https://issues.apache.org/jira/browse/HBASE-6260
> Project: HBase
> Issue Type: Task
> Components: master, zookeeper
> Affects Versions: 0.96.0
> Reporter: Gregory Chanan
> Assignee: Gregory Chanan
> Priority: Blocker
> Attachments: HBASE-6260-addendum2.patch, HBASE-6260-addendum2.patch, HBASE-6260-addendum.patch, HBASE-6260.patch, HBASE-6260-v2.patch
>
>
> See: https://issues.apache.org/jira/browse/HBASE-5953?focusedCommentId=13270200&page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel#comment-13270200
> And: https://issues.apache.org/jira/browse/HBASE-5630?focusedCommentId=13399225&page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel#comment-13399225
> In short, we need to move the balancer state to ZK so that it won't have to be restarted if the master dies.
--
This message is automatically generated by JIRA.
If you think it was sent incorrectly, please contact your JIRA administrators
For more information on JIRA, see: http://www.atlassian.com/software/jira
[jira] [Commented] (HBASE-6260) balancer state should be stored in
ZK
Posted by "Hadoop QA (JIRA)" <ji...@apache.org>.
[ https://issues.apache.org/jira/browse/HBASE-6260?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13455350#comment-13455350 ]
Hadoop QA commented on HBASE-6260:
----------------------------------
-1 overall. Here are the results of testing the latest attachment
http://issues.apache.org/jira/secure/attachment/12545054/HBASE-6260-v2.patch
against trunk revision .
+1 @author. The patch does not contain any @author tags.
+1 tests included. The patch appears to include 3 new or modified tests.
+1 hadoop2.0. The patch compiles against the hadoop 2.0 profile.
+1 javadoc. The javadoc tool did not generate any warning messages.
-1 javac. The patch appears to cause mvn compile goal to fail.
-1 findbugs. The patch appears to cause Findbugs (version 1.3.9) to fail.
+1 release audit. The applied patch does not increase the total number of release audit warnings.
-1 core tests. The patch failed these unit tests:
org.apache.hadoop.hbase.io.hfile.TestForceCacheImportantBlocks
org.apache.hadoop.hbase.replication.TestReplication
Test results: https://builds.apache.org/job/PreCommit-HBASE-Build/2859//testReport/
Console output: https://builds.apache.org/job/PreCommit-HBASE-Build/2859//console
This message is automatically generated.
> balancer state should be stored in ZK
> -------------------------------------
>
> Key: HBASE-6260
> URL: https://issues.apache.org/jira/browse/HBASE-6260
> Project: HBase
> Issue Type: Task
> Components: master, zookeeper
> Affects Versions: 0.96.0
> Reporter: Gregory Chanan
> Assignee: Gregory Chanan
> Priority: Blocker
> Attachments: HBASE-6260.patch, HBASE-6260-v2.patch
>
>
> See: https://issues.apache.org/jira/browse/HBASE-5953?focusedCommentId=13270200&page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel#comment-13270200
> And: https://issues.apache.org/jira/browse/HBASE-5630?focusedCommentId=13399225&page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel#comment-13399225
> In short, we need to move the balancer state to ZK so that it won't have to be restarted if the master dies.
--
This message is automatically generated by JIRA.
If you think it was sent incorrectly, please contact your JIRA administrators
For more information on JIRA, see: http://www.atlassian.com/software/jira
[jira] [Reopened] (HBASE-6260) balancer state should be stored in
ZK
Posted by "Gregory Chanan (JIRA)" <ji...@apache.org>.
[ https://issues.apache.org/jira/browse/HBASE-6260?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ]
Gregory Chanan reopened HBASE-6260:
-----------------------------------
> balancer state should be stored in ZK
> -------------------------------------
>
> Key: HBASE-6260
> URL: https://issues.apache.org/jira/browse/HBASE-6260
> Project: HBase
> Issue Type: Task
> Components: master, zookeeper
> Affects Versions: 0.96.0
> Reporter: Gregory Chanan
> Assignee: Gregory Chanan
> Priority: Blocker
> Attachments: HBASE-6260-addendum2.patch, HBASE-6260-addendum2.patch, HBASE-6260-addendum.patch, HBASE-6260.patch, HBASE-6260-v2.patch
>
>
> See: https://issues.apache.org/jira/browse/HBASE-5953?focusedCommentId=13270200&page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel#comment-13270200
> And: https://issues.apache.org/jira/browse/HBASE-5630?focusedCommentId=13399225&page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel#comment-13399225
> In short, we need to move the balancer state to ZK so that it won't have to be restarted if the master dies.
--
This message is automatically generated by JIRA.
If you think it was sent incorrectly, please contact your JIRA administrators
For more information on JIRA, see: http://www.atlassian.com/software/jira
[jira] [Updated] (HBASE-6260) balancer state should be stored in ZK
Posted by "Gregory Chanan (JIRA)" <ji...@apache.org>.
[ https://issues.apache.org/jira/browse/HBASE-6260?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ]
Gregory Chanan updated HBASE-6260:
----------------------------------
Attachment: HBASE-6260-addendum2.patch
> balancer state should be stored in ZK
> -------------------------------------
>
> Key: HBASE-6260
> URL: https://issues.apache.org/jira/browse/HBASE-6260
> Project: HBase
> Issue Type: Task
> Components: master, zookeeper
> Affects Versions: 0.96.0
> Reporter: Gregory Chanan
> Assignee: Gregory Chanan
> Priority: Blocker
> Attachments: HBASE-6260-addendum2.patch, HBASE-6260-addendum.patch, HBASE-6260.patch, HBASE-6260-v2.patch
>
>
> See: https://issues.apache.org/jira/browse/HBASE-5953?focusedCommentId=13270200&page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel#comment-13270200
> And: https://issues.apache.org/jira/browse/HBASE-5630?focusedCommentId=13399225&page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel#comment-13399225
> In short, we need to move the balancer state to ZK so that it won't have to be restarted if the master dies.
--
This message is automatically generated by JIRA.
If you think it was sent incorrectly, please contact your JIRA administrators
For more information on JIRA, see: http://www.atlassian.com/software/jira
[jira] [Updated] (HBASE-6260) balancer state should be stored in ZK
Posted by "Gregory Chanan (JIRA)" <ji...@apache.org>.
[ https://issues.apache.org/jira/browse/HBASE-6260?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ]
Gregory Chanan updated HBASE-6260:
----------------------------------
Attachment: HBASE-6260-addendum.patch
@Ted:
attached an addendum. Let me know what you think.
> balancer state should be stored in ZK
> -------------------------------------
>
> Key: HBASE-6260
> URL: https://issues.apache.org/jira/browse/HBASE-6260
> Project: HBase
> Issue Type: Task
> Components: master, zookeeper
> Affects Versions: 0.96.0
> Reporter: Gregory Chanan
> Assignee: Gregory Chanan
> Priority: Blocker
> Attachments: HBASE-6260-addendum.patch, HBASE-6260.patch, HBASE-6260-v2.patch
>
>
> See: https://issues.apache.org/jira/browse/HBASE-5953?focusedCommentId=13270200&page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel#comment-13270200
> And: https://issues.apache.org/jira/browse/HBASE-5630?focusedCommentId=13399225&page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel#comment-13399225
> In short, we need to move the balancer state to ZK so that it won't have to be restarted if the master dies.
--
This message is automatically generated by JIRA.
If you think it was sent incorrectly, please contact your JIRA administrators
For more information on JIRA, see: http://www.atlassian.com/software/jira
[jira] [Commented] (HBASE-6260) balancer state should be stored in
ZK
Posted by "stack (JIRA)" <ji...@apache.org>.
[ https://issues.apache.org/jira/browse/HBASE-6260?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13452685#comment-13452685 ]
stack commented on HBASE-6260:
------------------------------
+1 on patch.
ClientProtos.java change looks gratuitous. Leave it out of patch on commit?
Do you think we need to add the PB_MAGIC on the front of the data we write to zk? Its only a boolean so wouldn't be the end of the world if missing. Most of the other pb serializations to pb seem to have it though?
> balancer state should be stored in ZK
> -------------------------------------
>
> Key: HBASE-6260
> URL: https://issues.apache.org/jira/browse/HBASE-6260
> Project: HBase
> Issue Type: Task
> Components: master, zookeeper
> Affects Versions: 0.96.0
> Reporter: Gregory Chanan
> Assignee: Gregory Chanan
> Priority: Blocker
> Attachments: HBASE-6260.patch
>
>
> See: https://issues.apache.org/jira/browse/HBASE-5953?focusedCommentId=13270200&page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel#comment-13270200
> And: https://issues.apache.org/jira/browse/HBASE-5630?focusedCommentId=13399225&page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel#comment-13399225
> In short, we need to move the balancer state to ZK so that it won't have to be restarted if the master dies.
--
This message is automatically generated by JIRA.
If you think it was sent incorrectly, please contact your JIRA administrators
For more information on JIRA, see: http://www.atlassian.com/software/jira
[jira] [Updated] (HBASE-6260) balancer state should be stored in ZK
Posted by "Gregory Chanan (JIRA)" <ji...@apache.org>.
[ https://issues.apache.org/jira/browse/HBASE-6260?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ]
Gregory Chanan updated HBASE-6260:
----------------------------------
Resolution: Fixed
Status: Resolved (was: Patch Available)
Thanks for the review, stack. Committed to trunk.
> balancer state should be stored in ZK
> -------------------------------------
>
> Key: HBASE-6260
> URL: https://issues.apache.org/jira/browse/HBASE-6260
> Project: HBase
> Issue Type: Task
> Components: master, zookeeper
> Affects Versions: 0.96.0
> Reporter: Gregory Chanan
> Assignee: Gregory Chanan
> Priority: Blocker
> Attachments: HBASE-6260.patch, HBASE-6260-v2.patch
>
>
> See: https://issues.apache.org/jira/browse/HBASE-5953?focusedCommentId=13270200&page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel#comment-13270200
> And: https://issues.apache.org/jira/browse/HBASE-5630?focusedCommentId=13399225&page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel#comment-13399225
> In short, we need to move the balancer state to ZK so that it won't have to be restarted if the master dies.
--
This message is automatically generated by JIRA.
If you think it was sent incorrectly, please contact your JIRA administrators
For more information on JIRA, see: http://www.atlassian.com/software/jira
[jira] [Commented] (HBASE-6260) balancer state should be stored in
ZK
Posted by "Jonathan Hsieh (JIRA)" <ji...@apache.org>.
[ https://issues.apache.org/jira/browse/HBASE-6260?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13399683#comment-13399683 ]
Jonathan Hsieh commented on HBASE-6260:
---------------------------------------
While we are at it, we should add a method that only queries and returns the current state of the balancer. (and a new command in the shell). Currently it seems like we have to toggle it to the old value and then toggle it back which is annoying and potentially race prone.
> balancer state should be stored in ZK
> -------------------------------------
>
> Key: HBASE-6260
> URL: https://issues.apache.org/jira/browse/HBASE-6260
> Project: HBase
> Issue Type: Task
> Components: master, zookeeper
> Affects Versions: 0.96.0
> Reporter: Gregory Chanan
> Assignee: Gregory Chanan
>
> See: https://issues.apache.org/jira/browse/HBASE-5953?focusedCommentId=13270200&page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel#comment-13270200
> And: https://issues.apache.org/jira/browse/HBASE-5630?focusedCommentId=13399225&page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel#comment-13399225
> In short, we need to move the balancer state to ZK so that it won't have to be restarted if the master dies.
--
This message is automatically generated by JIRA.
If you think it was sent incorrectly, please contact your JIRA administrators: https://issues.apache.org/jira/secure/ContactAdministrators!default.jspa
For more information on JIRA, see: http://www.atlassian.com/software/jira
[jira] [Commented] (HBASE-6260) balancer state should be stored in
ZK
Posted by "Gregory Chanan (JIRA)" <ji...@apache.org>.
[ https://issues.apache.org/jira/browse/HBASE-6260?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13456183#comment-13456183 ]
Gregory Chanan commented on HBASE-6260:
---------------------------------------
HadoopQA hung in a couple tests:
{noformat}
dev-support/findHangingTest.sh https://builds.apache.org/job/PreCommit-HBASE-Build/2870//console
% Total % Received % Xferd Average Speed Time Time Time Current
Dload Upload Total Spent Left Speed
100 159k 0 159k 0 0 563k 0 --:--:-- --:--:-- --:--:-- 1413k
Hanging test: Running org.apache.hadoop.hbase.util.hbck.TestOfflineMetaRebuildOverlap
Hanging test: Running org.apache.hadoop.hbase.util.TestHBaseFsck
{noformat}
Ran the unit tests locally and they passed:
{noformat}
Running org.apache.hadoop.hbase.util.hbck.TestOfflineMetaRebuildOverlap
Tests run: 1, Failures: 0, Errors: 0, Skipped: 0, Time elapsed: 18.809 sec
{noformat}
{noformat}
Running org.apache.hadoop.hbase.util.TestHBaseFsck
Tests run: 27, Failures: 0, Errors: 0, Skipped: 0, Time elapsed: 210.967 sec
{noformat}
> balancer state should be stored in ZK
> -------------------------------------
>
> Key: HBASE-6260
> URL: https://issues.apache.org/jira/browse/HBASE-6260
> Project: HBase
> Issue Type: Task
> Components: master, zookeeper
> Affects Versions: 0.96.0
> Reporter: Gregory Chanan
> Assignee: Gregory Chanan
> Priority: Blocker
> Attachments: 6260-addendum-3.txt, HBASE-6260-addendum2.patch, HBASE-6260-addendum2.patch, HBASE-6260-addendum.patch, HBASE-6260.patch, HBASE-6260-v2.patch
>
>
> See: https://issues.apache.org/jira/browse/HBASE-5953?focusedCommentId=13270200&page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel#comment-13270200
> And: https://issues.apache.org/jira/browse/HBASE-5630?focusedCommentId=13399225&page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel#comment-13399225
> In short, we need to move the balancer state to ZK so that it won't have to be restarted if the master dies.
--
This message is automatically generated by JIRA.
If you think it was sent incorrectly, please contact your JIRA administrators
For more information on JIRA, see: http://www.atlassian.com/software/jira