You are viewing a plain text version of this content. The canonical link for it is here.
Posted to common-dev@hadoop.apache.org by "David Bowen (JIRA)" <ji...@apache.org> on 2007/02/27 20:23:05 UTC
[jira] Created: (HADOOP-1048) Sort500 failing since counters patch
went in
Sort500 failing since counters patch went in
--------------------------------------------
Key: HADOOP-1048
URL: https://issues.apache.org/jira/browse/HADOOP-1048
Project: Hadoop
Issue Type: Bug
Components: mapred
Reporter: David Bowen
Nigel wrote:
We have a problem
Looks like counters patch is "breaking" things
Sort500 has been failing since the patch went in
Looking at the JT logs, there are 0 "Call queue overflow" message before the patch went in and 15000+ after the patch went in
Looks like the counters stuff is overwhelming the JT
--
This message is automatically generated by JIRA.
-
You can reply to this email to add a comment to the issue online.
[jira] Resolved: (HADOOP-1048) Sort500 failing since counters patch
went in
Posted by "David Bowen (JIRA)" <ji...@apache.org>.
[ https://issues.apache.org/jira/browse/HADOOP-1048?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ]
David Bowen resolved HADOOP-1048.
---------------------------------
Resolution: Duplicate
This issue is now being tracked as part of Hadoop-1041.
> Sort500 failing since counters patch went in
> --------------------------------------------
>
> Key: HADOOP-1048
> URL: https://issues.apache.org/jira/browse/HADOOP-1048
> Project: Hadoop
> Issue Type: Bug
> Components: mapred
> Reporter: David Bowen
> Assigned To: David Bowen
>
> Nigel wrote:
> We have a problem
> Looks like counters patch is "breaking" things
> Sort500 has been failing since the patch went in
> Looking at the JT logs, there are 0 "Call queue overflow" message before the patch went in and 15000+ after the patch went in
> Looks like the counters stuff is overwhelming the JT
--
This message is automatically generated by JIRA.
-
You can reply to this email to add a comment to the issue online.
[jira] Commented: (HADOOP-1048) Sort500 failing since counters
patch went in
Posted by "Sameer Paranjpye (JIRA)" <ji...@apache.org>.
[ https://issues.apache.org/jira/browse/HADOOP-1048?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel#action_12477035 ]
Sameer Paranjpye commented on HADOOP-1048:
------------------------------------------
Yes, looks like HADOOP-1048 is subsumed by HADOOP-1041
> Sort500 failing since counters patch went in
> --------------------------------------------
>
> Key: HADOOP-1048
> URL: https://issues.apache.org/jira/browse/HADOOP-1048
> Project: Hadoop
> Issue Type: Bug
> Components: mapred
> Reporter: David Bowen
> Assigned To: David Bowen
>
> Nigel wrote:
> We have a problem
> Looks like counters patch is "breaking" things
> Sort500 has been failing since the patch went in
> Looking at the JT logs, there are 0 "Call queue overflow" message before the patch went in and 15000+ after the patch went in
> Looks like the counters stuff is overwhelming the JT
--
This message is automatically generated by JIRA.
-
You can reply to this email to add a comment to the issue online.
[jira] Assigned: (HADOOP-1048) Sort500 failing since counters patch
went in
Posted by "David Bowen (JIRA)" <ji...@apache.org>.
[ https://issues.apache.org/jira/browse/HADOOP-1048?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ]
David Bowen reassigned HADOOP-1048:
-----------------------------------
Assignee: David Bowen
> Sort500 failing since counters patch went in
> --------------------------------------------
>
> Key: HADOOP-1048
> URL: https://issues.apache.org/jira/browse/HADOOP-1048
> Project: Hadoop
> Issue Type: Bug
> Components: mapred
> Reporter: David Bowen
> Assigned To: David Bowen
>
> Nigel wrote:
> We have a problem
> Looks like counters patch is "breaking" things
> Sort500 has been failing since the patch went in
> Looking at the JT logs, there are 0 "Call queue overflow" message before the patch went in and 15000+ after the patch went in
> Looks like the counters stuff is overwhelming the JT
--
This message is automatically generated by JIRA.
-
You can reply to this email to add a comment to the issue online.
[jira] Commented: (HADOOP-1048) Sort500 failing since counters
patch went in
Posted by "Doug Cutting (JIRA)" <ji...@apache.org>.
[ https://issues.apache.org/jira/browse/HADOOP-1048?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel#action_12476375 ]
Doug Cutting commented on HADOOP-1048:
--------------------------------------
Another possible optimization in JobInProgress#updateTaskStatus might be to keep, for each task, its previous counter values, then compute the difference between those and the new values, and update global counters incrementally rather than recomputing them from scratch each time. This might not actually reduce the amount of computation much, but it might greatly decrease the amount of allocated objects, and object allocations are considerably more expensive than arithmetic.
> Sort500 failing since counters patch went in
> --------------------------------------------
>
> Key: HADOOP-1048
> URL: https://issues.apache.org/jira/browse/HADOOP-1048
> Project: Hadoop
> Issue Type: Bug
> Components: mapred
> Reporter: David Bowen
>
> Nigel wrote:
> We have a problem
> Looks like counters patch is "breaking" things
> Sort500 has been failing since the patch went in
> Looking at the JT logs, there are 0 "Call queue overflow" message before the patch went in and 15000+ after the patch went in
> Looks like the counters stuff is overwhelming the JT
--
This message is automatically generated by JIRA.
-
You can reply to this email to add a comment to the issue online.
[jira] Commented: (HADOOP-1048) Sort500 failing since counters
patch went in
Posted by "David Bowen (JIRA)" <ji...@apache.org>.
[ https://issues.apache.org/jira/browse/HADOOP-1048?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel#action_12476358 ]
David Bowen commented on HADOOP-1048:
-------------------------------------
Currently tasks are reporting their progress (and now counters) every second. Is there any reason for this to be so frequent? Shall I try setting Task.PROGRESS_INTERVAL to 5 seconds instead?
> Sort500 failing since counters patch went in
> --------------------------------------------
>
> Key: HADOOP-1048
> URL: https://issues.apache.org/jira/browse/HADOOP-1048
> Project: Hadoop
> Issue Type: Bug
> Components: mapred
> Reporter: David Bowen
>
> Nigel wrote:
> We have a problem
> Looks like counters patch is "breaking" things
> Sort500 has been failing since the patch went in
> Looking at the JT logs, there are 0 "Call queue overflow" message before the patch went in and 15000+ after the patch went in
> Looks like the counters stuff is overwhelming the JT
--
This message is automatically generated by JIRA.
-
You can reply to this email to add a comment to the issue online.
[jira] Commented: (HADOOP-1048) Sort500 failing since counters
patch went in
Posted by "Doug Cutting (JIRA)" <ji...@apache.org>.
[ https://issues.apache.org/jira/browse/HADOOP-1048?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel#action_12476363 ]
Doug Cutting commented on HADOOP-1048:
--------------------------------------
It might also be interesting to profile the counter code. I wonder how much faster it would be if instead of a Map<String,Long> we used a Map<String,CounterValue> where CounterValue is defined as something like:
private class CounterValue { long value; }
Then incrementing the value would not require creating a new Long instance.
> Sort500 failing since counters patch went in
> --------------------------------------------
>
> Key: HADOOP-1048
> URL: https://issues.apache.org/jira/browse/HADOOP-1048
> Project: Hadoop
> Issue Type: Bug
> Components: mapred
> Reporter: David Bowen
>
> Nigel wrote:
> We have a problem
> Looks like counters patch is "breaking" things
> Sort500 has been failing since the patch went in
> Looking at the JT logs, there are 0 "Call queue overflow" message before the patch went in and 15000+ after the patch went in
> Looks like the counters stuff is overwhelming the JT
--
This message is automatically generated by JIRA.
-
You can reply to this email to add a comment to the issue online.
[jira] Commented: (HADOOP-1048) Sort500 failing since counters
patch went in
Posted by "David Bowen (JIRA)" <ji...@apache.org>.
[ https://issues.apache.org/jira/browse/HADOOP-1048?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel#action_12477032 ]
David Bowen commented on HADOOP-1048:
-------------------------------------
Yes, the per-job totals can be compted lazily. This is implemented in the patch attached to 1041.
These two Jiras are separate issues with the same code. Would it make sense to merge them (i.e. close this one as a dup)?
> Sort500 failing since counters patch went in
> --------------------------------------------
>
> Key: HADOOP-1048
> URL: https://issues.apache.org/jira/browse/HADOOP-1048
> Project: Hadoop
> Issue Type: Bug
> Components: mapred
> Reporter: David Bowen
> Assigned To: David Bowen
>
> Nigel wrote:
> We have a problem
> Looks like counters patch is "breaking" things
> Sort500 has been failing since the patch went in
> Looking at the JT logs, there are 0 "Call queue overflow" message before the patch went in and 15000+ after the patch went in
> Looks like the counters stuff is overwhelming the JT
--
This message is automatically generated by JIRA.
-
You can reply to this email to add a comment to the issue online.
[jira] Commented: (HADOOP-1048) Sort500 failing since counters
patch went in
Posted by "Sameer Paranjpye (JIRA)" <ji...@apache.org>.
[ https://issues.apache.org/jira/browse/HADOOP-1048?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel#action_12477030 ]
Sameer Paranjpye commented on HADOOP-1048:
------------------------------------------
Could the global counters be computed lazily, on demand, with JobInProgress#updateTaskStatus only updating local values?
> Sort500 failing since counters patch went in
> --------------------------------------------
>
> Key: HADOOP-1048
> URL: https://issues.apache.org/jira/browse/HADOOP-1048
> Project: Hadoop
> Issue Type: Bug
> Components: mapred
> Reporter: David Bowen
> Assigned To: David Bowen
>
> Nigel wrote:
> We have a problem
> Looks like counters patch is "breaking" things
> Sort500 has been failing since the patch went in
> Looking at the JT logs, there are 0 "Call queue overflow" message before the patch went in and 15000+ after the patch went in
> Looks like the counters stuff is overwhelming the JT
--
This message is automatically generated by JIRA.
-
You can reply to this email to add a comment to the issue online.
[jira] Commented: (HADOOP-1048) Sort500 failing since counters
patch went in
Posted by "Doug Cutting (JIRA)" <ji...@apache.org>.
[ https://issues.apache.org/jira/browse/HADOOP-1048?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel#action_12476331 ]
Doug Cutting commented on HADOOP-1048:
--------------------------------------
Can we cause counters to somehow be updated far less frequently?
> Sort500 failing since counters patch went in
> --------------------------------------------
>
> Key: HADOOP-1048
> URL: https://issues.apache.org/jira/browse/HADOOP-1048
> Project: Hadoop
> Issue Type: Bug
> Components: mapred
> Reporter: David Bowen
>
> Nigel wrote:
> We have a problem
> Looks like counters patch is "breaking" things
> Sort500 has been failing since the patch went in
> Looking at the JT logs, there are 0 "Call queue overflow" message before the patch went in and 15000+ after the patch went in
> Looks like the counters stuff is overwhelming the JT
--
This message is automatically generated by JIRA.
-
You can reply to this email to add a comment to the issue online.