You are viewing a plain text version of this content. The canonical link for it is here.
Posted to issues@hbase.apache.org by "Jean-Daniel Cryans (JIRA)" <ji...@apache.org> on 2010/11/10 00:12:23 UTC
[jira] Created: (HBASE-3210) HBASE-1921 for the new master
HBASE-1921 for the new master
-----------------------------
Key: HBASE-3210
URL: https://issues.apache.org/jira/browse/HBASE-3210
Project: HBase
Issue Type: Improvement
Reporter: Jean-Daniel Cryans
Priority: Critical
Fix For: 0.92.0
HBASE-1921 was lost when writing the new master code. I guess it's going to be much harder to implement now, but I think it's a critical feature to have considering the reasons that brought me do it in the old master. There's already a test in TestZooKeeper which has been disabled a while ago.
--
This message is automatically generated by JIRA.
-
You can reply to this email to add a comment to the issue online.
[jira] [Updated] (HBASE-3210) HBASE-1921 for the new master
Posted by "Subbu M Iyer (JIRA)" <ji...@apache.org>.
[ https://issues.apache.org/jira/browse/HBASE-3210?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ]
Subbu M Iyer updated HBASE-3210:
--------------------------------
Attachment: HBASE-3210-When_the_Master_s_session_times_out_and_there_s_only_one_cluster_is_wedged-2.patch
> HBASE-1921 for the new master
> -----------------------------
>
> Key: HBASE-3210
> URL: https://issues.apache.org/jira/browse/HBASE-3210
> Project: HBase
> Issue Type: Improvement
> Reporter: Jean-Daniel Cryans
> Priority: Critical
> Fix For: 0.92.0
>
> Attachments: HBASE-3210-When_the_Master_s_session_times_out_and_there_s_only_one,_cluster_is_wedged.patch, HBASE-3210-When_the_Master_s_session_times_out_and_there_s_only_one_cluster_is_wedged-2.patch
>
>
> HBASE-1921 was lost when writing the new master code. I guess it's going to be much harder to implement now, but I think it's a critical feature to have considering the reasons that brought me do it in the old master. There's already a test in TestZooKeeper which has been disabled a while ago.
--
This message is automatically generated by JIRA.
For more information on JIRA, see: http://www.atlassian.com/software/jira
[jira] [Commented] (HBASE-3210) HBASE-1921 for the new master
Posted by "stack (JIRA)" <ji...@apache.org>.
[ https://issues.apache.org/jira/browse/HBASE-3210?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13019693#comment-13019693 ]
stack commented on HBASE-3210:
------------------------------
Subbu: Your patch looks great (as does your reenabling of TZK). I'm up for committing it -- was going to run all tests first first though since a pretty significant change -- but your patch is 4x the time it needs to be since the bulk is formatting only changes. Would you mind resubmitting the patch absent the formatting changes. Try also to keep lines < 80. Good stuff Subbu.
> HBASE-1921 for the new master
> -----------------------------
>
> Key: HBASE-3210
> URL: https://issues.apache.org/jira/browse/HBASE-3210
> Project: HBase
> Issue Type: Improvement
> Reporter: Jean-Daniel Cryans
> Priority: Critical
> Fix For: 0.92.0
>
> Attachments: HBASE-3210-When_the_Master_s_session_times_out_and_there_s_only_one,_cluster_is_wedged.patch
>
>
> HBASE-1921 was lost when writing the new master code. I guess it's going to be much harder to implement now, but I think it's a critical feature to have considering the reasons that brought me do it in the old master. There's already a test in TestZooKeeper which has been disabled a while ago.
--
This message is automatically generated by JIRA.
For more information on JIRA, see: http://www.atlassian.com/software/jira
[jira] [Updated] (HBASE-3210) HBASE-1921 for the new master
Posted by "Subbu M Iyer (JIRA)" <ji...@apache.org>.
[ https://issues.apache.org/jira/browse/HBASE-3210?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ]
Subbu M Iyer updated HBASE-3210:
--------------------------------
Attachment: HBASE-3210-When_the_Master_s_session_times_out_and_there_s_only_one,_cluster_is_wedged.patch
> HBASE-1921 for the new master
> -----------------------------
>
> Key: HBASE-3210
> URL: https://issues.apache.org/jira/browse/HBASE-3210
> Project: HBase
> Issue Type: Improvement
> Reporter: Jean-Daniel Cryans
> Priority: Critical
> Fix For: 0.92.0
>
> Attachments: HBASE-3210-When_the_Master_s_session_times_out_and_there_s_only_one,_cluster_is_wedged.patch
>
>
> HBASE-1921 was lost when writing the new master code. I guess it's going to be much harder to implement now, but I think it's a critical feature to have considering the reasons that brought me do it in the old master. There's already a test in TestZooKeeper which has been disabled a while ago.
--
This message is automatically generated by JIRA.
For more information on JIRA, see: http://www.atlassian.com/software/jira
[jira] [Commented] (HBASE-3210) HBASE-1921 for the new master
Posted by "Jean-Daniel Cryans (JIRA)" <ji...@apache.org>.
[ https://issues.apache.org/jira/browse/HBASE-3210?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13022317#comment-13022317 ]
Jean-Daniel Cryans commented on HBASE-3210:
-------------------------------------------
Comments after first pass:
- becomeActiveMaster and recoverableFromZKSessionExpiry needs complete javadoc (the @return tag is empty)
- should recoverableFromZKSessionExpiry be called tryRecoveringExpiredSession?
- in recoverableFromZKSessionExpiry, you do a big if(becomeActiveMaster()) and then return true, or false. What we usually prefer for readability is to instead do something like:
{code}
if (!becomeActiveMaster()) {
return false;
}
// Initialize ZK based trackers since we now have a new ZK session.
initilizeZKBasedSystemTrackers();
// Update in-memory strutures to reflect our earlier Root/Meta assignment.
assignRootAndMeta();
// process RIT if any
this.assignmentManager.processRegionsInTransition();
return true;
{code}
- Instead of:
{code}
return recoverableFromZKSessionExpiry() ? false : true;
{code}
do
{code}
return !recoverableFromZKSessionExpiry();
{code}
- abortNow in abortNow() is set multiple times to true. Review how you use that variable.
- There's a typo in initilizeZKBasedSystemTrackers
> HBASE-1921 for the new master
> -----------------------------
>
> Key: HBASE-3210
> URL: https://issues.apache.org/jira/browse/HBASE-3210
> Project: HBase
> Issue Type: Improvement
> Reporter: Jean-Daniel Cryans
> Priority: Critical
> Fix For: 0.92.0
>
> Attachments: HBASE-3210-When_the_Master_s_session_times_out_and_there_s_only_one,_cluster_is_wedged.patch, HBASE-3210-When_the_Master_s_session_times_out_and_there_s_only_one_cluster_is_wedged-2.patch
>
>
> HBASE-1921 was lost when writing the new master code. I guess it's going to be much harder to implement now, but I think it's a critical feature to have considering the reasons that brought me do it in the old master. There's already a test in TestZooKeeper which has been disabled a while ago.
--
This message is automatically generated by JIRA.
For more information on JIRA, see: http://www.atlassian.com/software/jira
[jira] [Commented] (HBASE-3210) HBASE-1921 for the new master
Posted by "stack (JIRA)" <ji...@apache.org>.
[ https://issues.apache.org/jira/browse/HBASE-3210?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13022296#comment-13022296 ]
stack commented on HBASE-3210:
------------------------------
running unit tests to ensure this change breaks nothing else
> HBASE-1921 for the new master
> -----------------------------
>
> Key: HBASE-3210
> URL: https://issues.apache.org/jira/browse/HBASE-3210
> Project: HBase
> Issue Type: Improvement
> Reporter: Jean-Daniel Cryans
> Priority: Critical
> Fix For: 0.92.0
>
> Attachments: HBASE-3210-When_the_Master_s_session_times_out_and_there_s_only_one,_cluster_is_wedged.patch, HBASE-3210-When_the_Master_s_session_times_out_and_there_s_only_one_cluster_is_wedged-2.patch
>
>
> HBASE-1921 was lost when writing the new master code. I guess it's going to be much harder to implement now, but I think it's a critical feature to have considering the reasons that brought me do it in the old master. There's already a test in TestZooKeeper which has been disabled a while ago.
--
This message is automatically generated by JIRA.
For more information on JIRA, see: http://www.atlassian.com/software/jira
[jira] [Commented] (HBASE-3210) HBASE-1921 for the new master
Posted by "Hudson (JIRA)" <ji...@apache.org>.
[ https://issues.apache.org/jira/browse/HBASE-3210?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13027255#comment-13027255 ]
Hudson commented on HBASE-3210:
-------------------------------
Integrated in HBase-TRUNK #1888 (See [https://builds.apache.org/hudson/job/HBase-TRUNK/1888/])
> HBASE-1921 for the new master
> -----------------------------
>
> Key: HBASE-3210
> URL: https://issues.apache.org/jira/browse/HBASE-3210
> Project: HBase
> Issue Type: Improvement
> Reporter: Jean-Daniel Cryans
> Assignee: Subbu M Iyer
> Priority: Critical
> Fix For: 0.92.0
>
> Attachments: HBASE-3210-When_the_Master_s_session_times_out_and_there_s_only_one,_cluster_is_wedged.patch, HBASE-3210-When_the_Master_s_session_times_out_and_there_s_only_one_cluster_is_wedged-2.patch, HBASE-3210-When_the_Master_s_session_times_out_and_there_s_only_one_cluster_is_wedged-3.patch
>
>
> HBASE-1921 was lost when writing the new master code. I guess it's going to be much harder to implement now, but I think it's a critical feature to have considering the reasons that brought me do it in the old master. There's already a test in TestZooKeeper which has been disabled a while ago.
--
This message is automatically generated by JIRA.
For more information on JIRA, see: http://www.atlassian.com/software/jira
[jira] [Updated] (HBASE-3210) HBASE-1921 for the new master
Posted by "Subbu M Iyer (JIRA)" <ji...@apache.org>.
[ https://issues.apache.org/jira/browse/HBASE-3210?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ]
Subbu M Iyer updated HBASE-3210:
--------------------------------
Attachment: HBASE-3210-When_the_Master_s_session_times_out_and_there_s_only_one_cluster_is_wedged-3.patch
> HBASE-1921 for the new master
> -----------------------------
>
> Key: HBASE-3210
> URL: https://issues.apache.org/jira/browse/HBASE-3210
> Project: HBase
> Issue Type: Improvement
> Reporter: Jean-Daniel Cryans
> Priority: Critical
> Fix For: 0.92.0
>
> Attachments: HBASE-3210-When_the_Master_s_session_times_out_and_there_s_only_one,_cluster_is_wedged.patch, HBASE-3210-When_the_Master_s_session_times_out_and_there_s_only_one_cluster_is_wedged-2.patch, HBASE-3210-When_the_Master_s_session_times_out_and_there_s_only_one_cluster_is_wedged-3.patch
>
>
> HBASE-1921 was lost when writing the new master code. I guess it's going to be much harder to implement now, but I think it's a critical feature to have considering the reasons that brought me do it in the old master. There's already a test in TestZooKeeper which has been disabled a while ago.
--
This message is automatically generated by JIRA.
For more information on JIRA, see: http://www.atlassian.com/software/jira
[jira] [Commented] (HBASE-3210) HBASE-1921 for the new master
Posted by "Subbu M Iyer (JIRA)" <ji...@apache.org>.
[ https://issues.apache.org/jira/browse/HBASE-3210?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13015089#comment-13015089 ]
Subbu M Iyer commented on HBASE-3210:
-------------------------------------
First draft of my patch for review.
Here is what is being done now:
1. When the primary master's Abort is triggered from ZK Node listener during a ZK session expiry event, we first try to see if we can restore the ZK session. We ignore the abort trigger and continue working as primary master, if we can successfully restore the ZK session.
2. A successful ZK session recovery involves the following.
a. Create a ZK Session
b. Try becoming the primary master again. (so that we don't step onto secondary master's toes)
c. Initialize all ZK based trackers. This includes the AssignmentManager, CatalogTracker,
RegionServerTracker and ClusterSTatusTracker.
d. Assign Root and Meta. (We just ensure that our local memory structures are correctly updated to reflect our earlier Root/Meta assignments)
e. Process RIT if any, that came in during our blackout.
3. Refactored the Master startup logic so that we can reuse them during a master session recovery attempt.
> HBASE-1921 for the new master
> -----------------------------
>
> Key: HBASE-3210
> URL: https://issues.apache.org/jira/browse/HBASE-3210
> Project: HBase
> Issue Type: Improvement
> Reporter: Jean-Daniel Cryans
> Priority: Critical
> Fix For: 0.92.0
>
> Attachments: HBASE-3210-When_the_Master_s_session_times_out_and_there_s_only_one,_cluster_is_wedged.patch
>
>
> HBASE-1921 was lost when writing the new master code. I guess it's going to be much harder to implement now, but I think it's a critical feature to have considering the reasons that brought me do it in the old master. There's already a test in TestZooKeeper which has been disabled a while ago.
--
This message is automatically generated by JIRA.
For more information on JIRA, see: http://www.atlassian.com/software/jira
[jira] [Commented] (HBASE-3210) HBASE-1921 for the new master
Posted by "Subbu M Iyer (JIRA)" <ji...@apache.org>.
[ https://issues.apache.org/jira/browse/HBASE-3210?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13021921#comment-13021921 ]
Subbu M Iyer commented on HBASE-3210:
-------------------------------------
Sure deal. will do it ASAP.
Thanks Stack for the review.
> HBASE-1921 for the new master
> -----------------------------
>
> Key: HBASE-3210
> URL: https://issues.apache.org/jira/browse/HBASE-3210
> Project: HBase
> Issue Type: Improvement
> Reporter: Jean-Daniel Cryans
> Priority: Critical
> Fix For: 0.92.0
>
> Attachments: HBASE-3210-When_the_Master_s_session_times_out_and_there_s_only_one,_cluster_is_wedged.patch
>
>
> HBASE-1921 was lost when writing the new master code. I guess it's going to be much harder to implement now, but I think it's a critical feature to have considering the reasons that brought me do it in the old master. There's already a test in TestZooKeeper which has been disabled a while ago.
--
This message is automatically generated by JIRA.
For more information on JIRA, see: http://www.atlassian.com/software/jira
[jira] [Commented] (HBASE-3210) HBASE-1921 for the new master
Posted by "Subbu M Iyer (JIRA)" <ji...@apache.org>.
[ https://issues.apache.org/jira/browse/HBASE-3210?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13022605#comment-13022605 ]
Subbu M Iyer commented on HBASE-3210:
-------------------------------------
Attached latest patch that addresses JD's comments/suggestions.
Thanks Stack for running the tests and I did it too.
thanks again.
> HBASE-1921 for the new master
> -----------------------------
>
> Key: HBASE-3210
> URL: https://issues.apache.org/jira/browse/HBASE-3210
> Project: HBase
> Issue Type: Improvement
> Reporter: Jean-Daniel Cryans
> Priority: Critical
> Fix For: 0.92.0
>
> Attachments: HBASE-3210-When_the_Master_s_session_times_out_and_there_s_only_one,_cluster_is_wedged.patch, HBASE-3210-When_the_Master_s_session_times_out_and_there_s_only_one_cluster_is_wedged-2.patch, HBASE-3210-When_the_Master_s_session_times_out_and_there_s_only_one_cluster_is_wedged-3.patch
>
>
> HBASE-1921 was lost when writing the new master code. I guess it's going to be much harder to implement now, but I think it's a critical feature to have considering the reasons that brought me do it in the old master. There's already a test in TestZooKeeper which has been disabled a while ago.
--
This message is automatically generated by JIRA.
For more information on JIRA, see: http://www.atlassian.com/software/jira
[jira] [Commented] (HBASE-3210) HBASE-1921 for the new master
Posted by "Subbu M Iyer (JIRA)" <ji...@apache.org>.
[ https://issues.apache.org/jira/browse/HBASE-3210?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13015090#comment-13015090 ]
Subbu M Iyer commented on HBASE-3210:
-------------------------------------
JD/Stack: Can one of your review this closely as it involves some minor refactoring to master startup? Also, JD worked on HBASE-1921 earlier and has the full context.
> HBASE-1921 for the new master
> -----------------------------
>
> Key: HBASE-3210
> URL: https://issues.apache.org/jira/browse/HBASE-3210
> Project: HBase
> Issue Type: Improvement
> Reporter: Jean-Daniel Cryans
> Priority: Critical
> Fix For: 0.92.0
>
> Attachments: HBASE-3210-When_the_Master_s_session_times_out_and_there_s_only_one,_cluster_is_wedged.patch
>
>
> HBASE-1921 was lost when writing the new master code. I guess it's going to be much harder to implement now, but I think it's a critical feature to have considering the reasons that brought me do it in the old master. There's already a test in TestZooKeeper which has been disabled a while ago.
--
This message is automatically generated by JIRA.
For more information on JIRA, see: http://www.atlassian.com/software/jira
[jira] [Resolved] (HBASE-3210) HBASE-1921 for the new master
Posted by "stack (JIRA)" <ji...@apache.org>.
[ https://issues.apache.org/jira/browse/HBASE-3210?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ]
stack resolved HBASE-3210.
--------------------------
Resolution: Fixed
Assignee: Subbu M Iyer
Hadoop Flags: [Reviewed]
Committed to TRUNK. Thank you for the patch Subbu.
> HBASE-1921 for the new master
> -----------------------------
>
> Key: HBASE-3210
> URL: https://issues.apache.org/jira/browse/HBASE-3210
> Project: HBase
> Issue Type: Improvement
> Reporter: Jean-Daniel Cryans
> Assignee: Subbu M Iyer
> Priority: Critical
> Fix For: 0.92.0
>
> Attachments: HBASE-3210-When_the_Master_s_session_times_out_and_there_s_only_one,_cluster_is_wedged.patch, HBASE-3210-When_the_Master_s_session_times_out_and_there_s_only_one_cluster_is_wedged-2.patch, HBASE-3210-When_the_Master_s_session_times_out_and_there_s_only_one_cluster_is_wedged-3.patch
>
>
> HBASE-1921 was lost when writing the new master code. I guess it's going to be much harder to implement now, but I think it's a critical feature to have considering the reasons that brought me do it in the old master. There's already a test in TestZooKeeper which has been disabled a while ago.
--
This message is automatically generated by JIRA.
For more information on JIRA, see: http://www.atlassian.com/software/jira
[jira] [Commented] (HBASE-3210) HBASE-1921 for the new master
Posted by "Subbu M Iyer (JIRA)" <ji...@apache.org>.
[ https://issues.apache.org/jira/browse/HBASE-3210?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13015091#comment-13015091 ]
Subbu M Iyer commented on HBASE-3210:
-------------------------------------
Also, I reenabled the disabled test in TestZooKeeper that tests for master session expiry.
All the tests are passing on my local.
> HBASE-1921 for the new master
> -----------------------------
>
> Key: HBASE-3210
> URL: https://issues.apache.org/jira/browse/HBASE-3210
> Project: HBase
> Issue Type: Improvement
> Reporter: Jean-Daniel Cryans
> Priority: Critical
> Fix For: 0.92.0
>
> Attachments: HBASE-3210-When_the_Master_s_session_times_out_and_there_s_only_one,_cluster_is_wedged.patch
>
>
> HBASE-1921 was lost when writing the new master code. I guess it's going to be much harder to implement now, but I think it's a critical feature to have considering the reasons that brought me do it in the old master. There's already a test in TestZooKeeper which has been disabled a while ago.
--
This message is automatically generated by JIRA.
For more information on JIRA, see: http://www.atlassian.com/software/jira
[jira] [Commented] (HBASE-3210) HBASE-1921 for the new master
Posted by "stack (JIRA)" <ji...@apache.org>.
[ https://issues.apache.org/jira/browse/HBASE-3210?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13022338#comment-13022338 ]
stack commented on HBASE-3210:
------------------------------
FYI, all tests pass for posted patch.
> HBASE-1921 for the new master
> -----------------------------
>
> Key: HBASE-3210
> URL: https://issues.apache.org/jira/browse/HBASE-3210
> Project: HBase
> Issue Type: Improvement
> Reporter: Jean-Daniel Cryans
> Priority: Critical
> Fix For: 0.92.0
>
> Attachments: HBASE-3210-When_the_Master_s_session_times_out_and_there_s_only_one,_cluster_is_wedged.patch, HBASE-3210-When_the_Master_s_session_times_out_and_there_s_only_one_cluster_is_wedged-2.patch
>
>
> HBASE-1921 was lost when writing the new master code. I guess it's going to be much harder to implement now, but I think it's a critical feature to have considering the reasons that brought me do it in the old master. There's already a test in TestZooKeeper which has been disabled a while ago.
--
This message is automatically generated by JIRA.
For more information on JIRA, see: http://www.atlassian.com/software/jira
[jira] [Commented] (HBASE-3210) HBASE-1921 for the new master
Posted by "Subbu M Iyer (JIRA)" <ji...@apache.org>.
[ https://issues.apache.org/jira/browse/HBASE-3210?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13022329#comment-13022329 ]
Subbu M Iyer commented on HBASE-3210:
-------------------------------------
JD:
thanks for your comments. Will address all your call outs.
> HBASE-1921 for the new master
> -----------------------------
>
> Key: HBASE-3210
> URL: https://issues.apache.org/jira/browse/HBASE-3210
> Project: HBase
> Issue Type: Improvement
> Reporter: Jean-Daniel Cryans
> Priority: Critical
> Fix For: 0.92.0
>
> Attachments: HBASE-3210-When_the_Master_s_session_times_out_and_there_s_only_one,_cluster_is_wedged.patch, HBASE-3210-When_the_Master_s_session_times_out_and_there_s_only_one_cluster_is_wedged-2.patch
>
>
> HBASE-1921 was lost when writing the new master code. I guess it's going to be much harder to implement now, but I think it's a critical feature to have considering the reasons that brought me do it in the old master. There's already a test in TestZooKeeper which has been disabled a while ago.
--
This message is automatically generated by JIRA.
For more information on JIRA, see: http://www.atlassian.com/software/jira
[jira] [Commented] (HBASE-3210) HBASE-1921 for the new master
Posted by "Subbu M Iyer (JIRA)" <ji...@apache.org>.
[ https://issues.apache.org/jira/browse/HBASE-3210?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13022235#comment-13022235 ]
Subbu M Iyer commented on HBASE-3210:
-------------------------------------
attaching the patch with corrected formatting and line breaks.
> HBASE-1921 for the new master
> -----------------------------
>
> Key: HBASE-3210
> URL: https://issues.apache.org/jira/browse/HBASE-3210
> Project: HBase
> Issue Type: Improvement
> Reporter: Jean-Daniel Cryans
> Priority: Critical
> Fix For: 0.92.0
>
> Attachments: HBASE-3210-When_the_Master_s_session_times_out_and_there_s_only_one,_cluster_is_wedged.patch, HBASE-3210-When_the_Master_s_session_times_out_and_there_s_only_one_cluster_is_wedged-2.patch
>
>
> HBASE-1921 was lost when writing the new master code. I guess it's going to be much harder to implement now, but I think it's a critical feature to have considering the reasons that brought me do it in the old master. There's already a test in TestZooKeeper which has been disabled a while ago.
--
This message is automatically generated by JIRA.
For more information on JIRA, see: http://www.atlassian.com/software/jira