You are viewing a plain text version of this content. The canonical link for it is here.
Posted to issues@hbase.apache.org by "Harsh J (JIRA)" <ji...@apache.org> on 2011/09/09 07:25:10 UTC
[jira] [Created] (HBASE-4359) Show dead RegionServer names in the
HMaster info page
Show dead RegionServer names in the HMaster info page
-----------------------------------------------------
Key: HBASE-4359
URL: https://issues.apache.org/jira/browse/HBASE-4359
Project: HBase
Issue Type: Improvement
Components: master
Affects Versions: 0.90.4
Reporter: Harsh J
Assignee: Harsh J
Priority: Trivial
Fix For: 0.94.0
Unlike other components of the cluster, like NameNode and JobTracker pages, the HMaster's info page does not show any data on dead region servers. While an RS is stateless being a good reason not to count dead nodes, I think having a list of dead nodes helps in cases where an administrator would want to find out which nodes are missing out on RS action (hey, everyone likes consistently spiking graphs! ;)).
Following HBASE-3580, I think it makes sense to have a list of already maintained dead nodes show up in the info UI.
--
This message is automatically generated by JIRA.
For more information on JIRA, see: http://www.atlassian.com/software/jira
[jira] [Updated] (HBASE-4359) Show dead RegionServer names in the
HMaster info page
Posted by "Harsh J (JIRA)" <ji...@apache.org>.
[ https://issues.apache.org/jira/browse/HBASE-4359?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ]
Harsh J updated HBASE-4359:
---------------------------
Status: Patch Available (was: Open)
> Show dead RegionServer names in the HMaster info page
> -----------------------------------------------------
>
> Key: HBASE-4359
> URL: https://issues.apache.org/jira/browse/HBASE-4359
> Project: HBase
> Issue Type: Improvement
> Components: master
> Affects Versions: 0.90.4
> Reporter: Harsh J
> Assignee: Harsh J
> Priority: Trivial
> Fix For: 0.94.0
>
> Attachments: HBASE-4359.r1.diff, HBase Master UI - Dead Servers.png
>
>
> Unlike other components of the cluster, like NameNode and JobTracker pages, the HMaster's info page does not show any data on dead region servers. While an RS is stateless being a good reason not to count dead nodes, I think having a list of dead nodes helps in cases where an administrator would want to find out which nodes are missing out on RS action (hey, everyone likes consistently spiking graphs! ;)).
> Following HBASE-3580, I think it makes sense to have a list of already maintained dead nodes show up in the info UI.
--
This message is automatically generated by JIRA.
For more information on JIRA, see: http://www.atlassian.com/software/jira
[jira] [Updated] (HBASE-4359) Show dead RegionServer names in the
HMaster info page
Posted by "stack (JIRA)" <ji...@apache.org>.
[ https://issues.apache.org/jira/browse/HBASE-4359?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ]
stack updated HBASE-4359:
-------------------------
Resolution: Fixed
Fix Version/s: (was: 0.94.0)
0.92.0
Hadoop Flags: [Reviewed]
Status: Resolved (was: Patch Available)
I tried it. We probably should not bother showing dead servers if none to show but otherwise looks good Harsh. This will do for now. Thanks for the patch. Applied to TRUNK.
> Show dead RegionServer names in the HMaster info page
> -----------------------------------------------------
>
> Key: HBASE-4359
> URL: https://issues.apache.org/jira/browse/HBASE-4359
> Project: HBase
> Issue Type: Improvement
> Components: master
> Affects Versions: 0.90.4
> Reporter: Harsh J
> Assignee: Harsh J
> Priority: Trivial
> Fix For: 0.92.0
>
> Attachments: HBASE-4359.r1.diff, HBASE-4359.r2.diff, HBase Master UI - Dead Servers (Yes, still dead).png, HBase Master UI - Dead Servers.png
>
>
> Unlike other components of the cluster, like NameNode and JobTracker pages, the HMaster's info page does not show any data on dead region servers. While an RS is stateless being a good reason not to count dead nodes, I think having a list of dead nodes helps in cases where an administrator would want to find out which nodes are missing out on RS action (hey, everyone likes consistently spiking graphs! ;)).
> Following HBASE-3580, I think it makes sense to have a list of already maintained dead nodes show up in the info UI.
--
This message is automatically generated by JIRA.
For more information on JIRA, see: http://www.atlassian.com/software/jira
[jira] [Commented] (HBASE-4359) Show dead RegionServer names in the
HMaster info page
Posted by "Hudson (JIRA)" <ji...@apache.org>.
[ https://issues.apache.org/jira/browse/HBASE-4359?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13104636#comment-13104636 ]
Hudson commented on HBASE-4359:
-------------------------------
Integrated in HBase-TRUNK #2208 (See [https://builds.apache.org/job/HBase-TRUNK/2208/])
HBASE-4359 Show dead RegionServer names in the HMaster info page
stack :
Files :
* /hbase/trunk/CHANGES.txt
* /hbase/trunk/src/main/jamon/org/apache/hbase/tmpl/master/MasterStatusTmpl.jamon
* /hbase/trunk/src/main/java/org/apache/hadoop/hbase/master/MasterStatusServlet.java
* /hbase/trunk/src/test/java/org/apache/hadoop/hbase/master/TestMasterStatusServlet.java
> Show dead RegionServer names in the HMaster info page
> -----------------------------------------------------
>
> Key: HBASE-4359
> URL: https://issues.apache.org/jira/browse/HBASE-4359
> Project: HBase
> Issue Type: Improvement
> Components: master
> Affects Versions: 0.90.4
> Reporter: Harsh J
> Assignee: Harsh J
> Priority: Trivial
> Fix For: 0.92.0
>
> Attachments: HBASE-4359.r1.diff, HBASE-4359.r2.diff, HBase Master UI - Dead Servers (Yes, still dead).png, HBase Master UI - Dead Servers.png
>
>
> Unlike other components of the cluster, like NameNode and JobTracker pages, the HMaster's info page does not show any data on dead region servers. While an RS is stateless being a good reason not to count dead nodes, I think having a list of dead nodes helps in cases where an administrator would want to find out which nodes are missing out on RS action (hey, everyone likes consistently spiking graphs! ;)).
> Following HBASE-3580, I think it makes sense to have a list of already maintained dead nodes show up in the info UI.
--
This message is automatically generated by JIRA.
For more information on JIRA, see: http://www.atlassian.com/software/jira
[jira] [Updated] (HBASE-4359) Show dead RegionServer names in the
HMaster info page
Posted by "Harsh J (JIRA)" <ji...@apache.org>.
[ https://issues.apache.org/jira/browse/HBASE-4359?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ]
Harsh J updated HBASE-4359:
---------------------------
Attachment: HBase Master UI - Dead Servers.png
HBASE-4359.r1.diff
Jamon+tests patch that adds in the improvement. Please review!
I ran the updated master status servlet test.
{code}
-> mvn -Dtest=TestMasterStatusServlet test
-------------------------------------------------------
T E S T S
-------------------------------------------------------
Running org.apache.hadoop.hbase.master.TestMasterStatusServlet
Tests run: 5, Failures: 0, Errors: 0, Skipped: 0, Time elapsed: 0.93 sec
Results :
Tests run: 5, Failures: 0, Errors: 0, Skipped: 0
{code}
I also ran the build manually via bin/start-hbase.sh with proper config and HDFS running. The screenshot shows the implementation.
Thanks in advance for reviews!
> Show dead RegionServer names in the HMaster info page
> -----------------------------------------------------
>
> Key: HBASE-4359
> URL: https://issues.apache.org/jira/browse/HBASE-4359
> Project: HBase
> Issue Type: Improvement
> Components: master
> Affects Versions: 0.90.4
> Reporter: Harsh J
> Assignee: Harsh J
> Priority: Trivial
> Fix For: 0.94.0
>
> Attachments: HBASE-4359.r1.diff, HBase Master UI - Dead Servers.png
>
>
> Unlike other components of the cluster, like NameNode and JobTracker pages, the HMaster's info page does not show any data on dead region servers. While an RS is stateless being a good reason not to count dead nodes, I think having a list of dead nodes helps in cases where an administrator would want to find out which nodes are missing out on RS action (hey, everyone likes consistently spiking graphs! ;)).
> Following HBASE-3580, I think it makes sense to have a list of already maintained dead nodes show up in the info UI.
--
This message is automatically generated by JIRA.
For more information on JIRA, see: http://www.atlassian.com/software/jira
[jira] [Updated] (HBASE-4359) Show dead RegionServer names in the
HMaster info page
Posted by "Harsh J (JIRA)" <ji...@apache.org>.
[ https://issues.apache.org/jira/browse/HBASE-4359?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ]
Harsh J updated HBASE-4359:
---------------------------
Attachment: HBase Master UI - Dead Servers (Yes, still dead).png
HBASE-4359.r2.diff
Newer patch after a chat with Todd. Made it more consistent with the online listing.
In brightest day, in darkest night, no test case shall escape my sight:
{code}
-------------------------------------------------------
T E S T S
-------------------------------------------------------
-------------------------------------------------------
T E S T S
-------------------------------------------------------
Running org.apache.hadoop.hbase.master.TestMasterStatusServlet
Tests run: 5, Failures: 0, Errors: 0, Skipped: 0, Time elapsed: 1.028 sec
Results :
Tests run: 5, Failures: 0, Errors: 0, Skipped: 0
{code}
Also upped a new UI image after manual testing.
> Show dead RegionServer names in the HMaster info page
> -----------------------------------------------------
>
> Key: HBASE-4359
> URL: https://issues.apache.org/jira/browse/HBASE-4359
> Project: HBase
> Issue Type: Improvement
> Components: master
> Affects Versions: 0.90.4
> Reporter: Harsh J
> Assignee: Harsh J
> Priority: Trivial
> Fix For: 0.94.0
>
> Attachments: HBASE-4359.r1.diff, HBASE-4359.r2.diff, HBase Master UI - Dead Servers (Yes, still dead).png, HBase Master UI - Dead Servers.png
>
>
> Unlike other components of the cluster, like NameNode and JobTracker pages, the HMaster's info page does not show any data on dead region servers. While an RS is stateless being a good reason not to count dead nodes, I think having a list of dead nodes helps in cases where an administrator would want to find out which nodes are missing out on RS action (hey, everyone likes consistently spiking graphs! ;)).
> Following HBASE-3580, I think it makes sense to have a list of already maintained dead nodes show up in the info UI.
--
This message is automatically generated by JIRA.
For more information on JIRA, see: http://www.atlassian.com/software/jira