You are viewing a plain text version of this content. The canonical link for it is here.
Posted to commits@cassandra.apache.org by "Jonathan Ellis (JIRA)" <ji...@apache.org> on 2011/02/16 07:11:57 UTC

[jira] Created: (CASSANDRA-2175) make key cache preheating use less memory

make key cache preheating use less memory
-----------------------------------------

                 Key: CASSANDRA-2175
                 URL: https://issues.apache.org/jira/browse/CASSANDRA-2175
             Project: Cassandra
          Issue Type: Improvement
    Affects Versions: 0.7.0 rc 3, 0.6.9
            Reporter: Jonathan Ellis
            Priority: Minor
             Fix For: 0.6.12, 0.7.3


CASSANDRA-1878 pre-heats the key cache post-compaction so latency doesn't suffer while warming the cache back up.  This can double the memory used temporarily; for a large key cache, this can have a substantial impact.

For now a boolean on/off is probably the best we can do.  With http://code.google.com/p/concurrentlinkedhashmap/issues/detail?id=21 though, we could say "preheat the hottest X keys."

-- 
This message is automatically generated by JIRA.
-
For more information on JIRA, see: http://www.atlassian.com/software/jira

        

[jira] Resolved: (CASSANDRA-2175) make key cache preheating use less memory

Posted by "Jonathan Ellis (JIRA)" <ji...@apache.org>.
     [ https://issues.apache.org/jira/browse/CASSANDRA-2175?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ]

Jonathan Ellis resolved CASSANDRA-2175.
---------------------------------------

    Resolution: Fixed
      Reviewer: brandon.williams
      Assignee: Jonathan Ellis

committed

> make key cache preheating use less memory
> -----------------------------------------
>
>                 Key: CASSANDRA-2175
>                 URL: https://issues.apache.org/jira/browse/CASSANDRA-2175
>             Project: Cassandra
>          Issue Type: Improvement
>    Affects Versions: 0.6.9, 0.7.0 rc 3
>            Reporter: Jonathan Ellis
>            Assignee: Jonathan Ellis
>            Priority: Minor
>             Fix For: 0.6.12, 0.7.3
>
>         Attachments: 2175-0.6.txt
>
>
> CASSANDRA-1878 pre-heats the key cache post-compaction so latency doesn't suffer while warming the cache back up.  This can double the memory used temporarily; for a large key cache, this can have a substantial impact.
> For now a boolean on/off is probably the best we can do.  With http://code.google.com/p/concurrentlinkedhashmap/issues/detail?id=21 though, we could say "preheat the hottest X keys."

-- 
This message is automatically generated by JIRA.
-
For more information on JIRA, see: http://www.atlassian.com/software/jira

        

[jira] Commented: (CASSANDRA-2175) make key cache preheating use less memory

Posted by "Brandon Williams (JIRA)" <ji...@apache.org>.
    [ https://issues.apache.org/jira/browse/CASSANDRA-2175?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=12996020#comment-12996020 ] 

Brandon Williams commented on CASSANDRA-2175:
---------------------------------------------

+1

> make key cache preheating use less memory
> -----------------------------------------
>
>                 Key: CASSANDRA-2175
>                 URL: https://issues.apache.org/jira/browse/CASSANDRA-2175
>             Project: Cassandra
>          Issue Type: Improvement
>    Affects Versions: 0.6.9, 0.7.0 rc 3
>            Reporter: Jonathan Ellis
>            Priority: Minor
>             Fix For: 0.6.12, 0.7.3
>
>         Attachments: 2175-0.6.txt
>
>
> CASSANDRA-1878 pre-heats the key cache post-compaction so latency doesn't suffer while warming the cache back up.  This can double the memory used temporarily; for a large key cache, this can have a substantial impact.
> For now a boolean on/off is probably the best we can do.  With http://code.google.com/p/concurrentlinkedhashmap/issues/detail?id=21 though, we could say "preheat the hottest X keys."

-- 
This message is automatically generated by JIRA.
-
For more information on JIRA, see: http://www.atlassian.com/software/jira

        

[jira] Updated: (CASSANDRA-2175) make key cache preheating use less memory

Posted by "Jonathan Ellis (JIRA)" <ji...@apache.org>.
     [ https://issues.apache.org/jira/browse/CASSANDRA-2175?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ]

Jonathan Ellis updated CASSANDRA-2175:
--------------------------------------

    Remaining Estimate: 4h
     Original Estimate: 4h

> make key cache preheating use less memory
> -----------------------------------------
>
>                 Key: CASSANDRA-2175
>                 URL: https://issues.apache.org/jira/browse/CASSANDRA-2175
>             Project: Cassandra
>          Issue Type: Improvement
>    Affects Versions: 0.6.9, 0.7.0 rc 3
>            Reporter: Jonathan Ellis
>            Assignee: Jonathan Ellis
>            Priority: Minor
>             Fix For: 0.6.12, 0.7.3
>
>         Attachments: 2175-0.6.txt
>
>   Original Estimate: 4h
>  Remaining Estimate: 4h
>
> CASSANDRA-1878 pre-heats the key cache post-compaction so latency doesn't suffer while warming the cache back up.  This can double the memory used temporarily; for a large key cache, this can have a substantial impact.
> For now a boolean on/off is probably the best we can do.  With http://code.google.com/p/concurrentlinkedhashmap/issues/detail?id=21 though, we could say "preheat the hottest X keys."

-- 
This message is automatically generated by JIRA.
-
For more information on JIRA, see: http://www.atlassian.com/software/jira

        

[jira] Updated: (CASSANDRA-2175) make key cache preheating use less memory

Posted by "Jonathan Ellis (JIRA)" <ji...@apache.org>.
     [ https://issues.apache.org/jira/browse/CASSANDRA-2175?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ]

Jonathan Ellis updated CASSANDRA-2175:
--------------------------------------

    Attachment: 2175-0.6.txt

adds -Dcompaction_preheat_key_cache (default false) for 0.6.  Will merge to 0.7 as a yaml property.

> make key cache preheating use less memory
> -----------------------------------------
>
>                 Key: CASSANDRA-2175
>                 URL: https://issues.apache.org/jira/browse/CASSANDRA-2175
>             Project: Cassandra
>          Issue Type: Improvement
>    Affects Versions: 0.6.9, 0.7.0 rc 3
>            Reporter: Jonathan Ellis
>            Priority: Minor
>             Fix For: 0.6.12, 0.7.3
>
>         Attachments: 2175-0.6.txt
>
>
> CASSANDRA-1878 pre-heats the key cache post-compaction so latency doesn't suffer while warming the cache back up.  This can double the memory used temporarily; for a large key cache, this can have a substantial impact.
> For now a boolean on/off is probably the best we can do.  With http://code.google.com/p/concurrentlinkedhashmap/issues/detail?id=21 though, we could say "preheat the hottest X keys."

-- 
This message is automatically generated by JIRA.
-
For more information on JIRA, see: http://www.atlassian.com/software/jira

        

[jira] Commented: (CASSANDRA-2175) make key cache preheating use less memory

Posted by "Hudson (JIRA)" <ji...@apache.org>.
    [ https://issues.apache.org/jira/browse/CASSANDRA-2175?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=12996107#comment-12996107 ] 

Hudson commented on CASSANDRA-2175:
-----------------------------------

Integrated in Cassandra-0.6 #60 (See [https://hudson.apache.org/hudson/job/Cassandra-0.6/60/])
    make key cache preheating default to false
patch by jbellis; reviewed by brandonwilliams for CASSANDRA-2175


> make key cache preheating use less memory
> -----------------------------------------
>
>                 Key: CASSANDRA-2175
>                 URL: https://issues.apache.org/jira/browse/CASSANDRA-2175
>             Project: Cassandra
>          Issue Type: Improvement
>    Affects Versions: 0.6.9, 0.7.0 rc 3
>            Reporter: Jonathan Ellis
>            Assignee: Jonathan Ellis
>            Priority: Minor
>             Fix For: 0.6.12, 0.7.3
>
>         Attachments: 2175-0.6.txt
>
>
> CASSANDRA-1878 pre-heats the key cache post-compaction so latency doesn't suffer while warming the cache back up.  This can double the memory used temporarily; for a large key cache, this can have a substantial impact.
> For now a boolean on/off is probably the best we can do.  With http://code.google.com/p/concurrentlinkedhashmap/issues/detail?id=21 though, we could say "preheat the hottest X keys."

-- 
This message is automatically generated by JIRA.
-
For more information on JIRA, see: http://www.atlassian.com/software/jira