You are viewing a plain text version of this content. The canonical link for it is here.
Posted to commits@cassandra.apache.org by "Yang Yang (JIRA)" <ji...@apache.org> on 2011/09/17 02:33:08 UTC
[jira] [Created] (CASSANDRA-3223) probably don't need to do full
copy to row cache after un-mmap() change
probably don't need to do full copy to row cache after un-mmap() change
-----------------------------------------------------------------------
Key: CASSANDRA-3223
URL: https://issues.apache.org/jira/browse/CASSANDRA-3223
Project: Cassandra
Issue Type: Bug
Reporter: Yang Yang
Priority: Minor
3179 changes from directly using the bytebuffer from mmap(), to copying that buffer,
CFS.cacheRow() https://github.com/apache/cassandra/blob/cassandra-1.0.0/src/java/org/apache/cassandra/db/ColumnFamilyStore.java line 1126
says it makes a deep copy exactly to prevent issues from unmmap().
maybe this deep copy is not needed now given 3179
if so, maybe slightly better performance in both speed and memory
--
This message is automatically generated by JIRA.
For more information on JIRA, see: http://www.atlassian.com/software/jira
[jira] [Updated] (CASSANDRA-3223) probably don't need to do full
copy to row cache after un-mmap() change
Posted by "Jonathan Ellis (JIRA)" <ji...@apache.org>.
[ https://issues.apache.org/jira/browse/CASSANDRA-3223?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ]
Jonathan Ellis updated CASSANDRA-3223:
--------------------------------------
Reviewer: jbellis
Component/s: Core
Affects Version/s: 1.0.0
Fix Version/s: 1.0.0
Assignee: Yang Yang
> probably don't need to do full copy to row cache after un-mmap() change
> -----------------------------------------------------------------------
>
> Key: CASSANDRA-3223
> URL: https://issues.apache.org/jira/browse/CASSANDRA-3223
> Project: Cassandra
> Issue Type: Bug
> Components: Core
> Affects Versions: 1.0.0
> Reporter: Yang Yang
> Assignee: Yang Yang
> Priority: Minor
> Fix For: 1.0.0
>
>
> 3179 changes from directly using the bytebuffer from mmap(), to copying that buffer,
> CFS.cacheRow() https://github.com/apache/cassandra/blob/cassandra-1.0.0/src/java/org/apache/cassandra/db/ColumnFamilyStore.java line 1126
> says it makes a deep copy exactly to prevent issues from unmmap().
> maybe this deep copy is not needed now given 3179
> if so, maybe slightly better performance in both speed and memory
--
This message is automatically generated by JIRA.
For more information on JIRA, see: http://www.atlassian.com/software/jira
[jira] [Commented] (CASSANDRA-3223) probably don't need to do full
copy to row cache after un-mmap() change
Posted by "Jonathan Ellis (JIRA)" <ji...@apache.org>.
[ https://issues.apache.org/jira/browse/CASSANDRA-3223?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13106998#comment-13106998 ]
Jonathan Ellis commented on CASSANDRA-3223:
-------------------------------------------
I think you're right. Care to submit a patch to remove the extra copy?
> probably don't need to do full copy to row cache after un-mmap() change
> -----------------------------------------------------------------------
>
> Key: CASSANDRA-3223
> URL: https://issues.apache.org/jira/browse/CASSANDRA-3223
> Project: Cassandra
> Issue Type: Bug
> Components: Core
> Affects Versions: 1.0.0
> Reporter: Yang Yang
> Priority: Minor
> Fix For: 1.0.0
>
>
> 3179 changes from directly using the bytebuffer from mmap(), to copying that buffer,
> CFS.cacheRow() https://github.com/apache/cassandra/blob/cassandra-1.0.0/src/java/org/apache/cassandra/db/ColumnFamilyStore.java line 1126
> says it makes a deep copy exactly to prevent issues from unmmap().
> maybe this deep copy is not needed now given 3179
> if so, maybe slightly better performance in both speed and memory
--
This message is automatically generated by JIRA.
For more information on JIRA, see: http://www.atlassian.com/software/jira
[jira] [Updated] (CASSANDRA-3223) probably don't need to do full
copy to row cache after un-mmap() change
Posted by "Yang Yang (JIRA)" <ji...@apache.org>.
[ https://issues.apache.org/jira/browse/CASSANDRA-3223?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ]
Yang Yang updated CASSANDRA-3223:
---------------------------------
Attachment: 0002-do-not-need-to-deep-copy-column-value-bytebuffer-now.patch
do not need to deep copy column value bytebuffer into row cache, now that we already do this copy in JIRA 3179
> probably don't need to do full copy to row cache after un-mmap() change
> -----------------------------------------------------------------------
>
> Key: CASSANDRA-3223
> URL: https://issues.apache.org/jira/browse/CASSANDRA-3223
> Project: Cassandra
> Issue Type: Bug
> Components: Core
> Affects Versions: 1.0.0
> Reporter: Yang Yang
> Assignee: Yang Yang
> Priority: Minor
> Fix For: 1.0.0
>
> Attachments: 0002-do-not-need-to-deep-copy-column-value-bytebuffer-now.patch
>
>
> 3179 changes from directly using the bytebuffer from mmap(), to copying that buffer,
> CFS.cacheRow() https://github.com/apache/cassandra/blob/cassandra-1.0.0/src/java/org/apache/cassandra/db/ColumnFamilyStore.java line 1126
> says it makes a deep copy exactly to prevent issues from unmmap().
> maybe this deep copy is not needed now given 3179
> if so, maybe slightly better performance in both speed and memory
--
This message is automatically generated by JIRA.
For more information on JIRA, see: http://www.atlassian.com/software/jira
[jira] [Commented] (CASSANDRA-3223) probably don't need to do full
copy to row cache after un-mmap() change
Posted by "Yang Yang (JIRA)" <ji...@apache.org>.
[ https://issues.apache.org/jira/browse/CASSANDRA-3223?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13107020#comment-13107020 ]
Yang Yang commented on CASSANDRA-3223:
--------------------------------------
sure
> probably don't need to do full copy to row cache after un-mmap() change
> -----------------------------------------------------------------------
>
> Key: CASSANDRA-3223
> URL: https://issues.apache.org/jira/browse/CASSANDRA-3223
> Project: Cassandra
> Issue Type: Bug
> Components: Core
> Affects Versions: 1.0.0
> Reporter: Yang Yang
> Assignee: Yang Yang
> Priority: Minor
> Fix For: 1.0.0
>
>
> 3179 changes from directly using the bytebuffer from mmap(), to copying that buffer,
> CFS.cacheRow() https://github.com/apache/cassandra/blob/cassandra-1.0.0/src/java/org/apache/cassandra/db/ColumnFamilyStore.java line 1126
> says it makes a deep copy exactly to prevent issues from unmmap().
> maybe this deep copy is not needed now given 3179
> if so, maybe slightly better performance in both speed and memory
--
This message is automatically generated by JIRA.
For more information on JIRA, see: http://www.atlassian.com/software/jira
[jira] [Resolved] (CASSANDRA-3223) probably don't need to do full
copy to row cache after un-mmap() change
Posted by "Jonathan Ellis (JIRA)" <ji...@apache.org>.
[ https://issues.apache.org/jira/browse/CASSANDRA-3223?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ]
Jonathan Ellis resolved CASSANDRA-3223.
---------------------------------------
Resolution: Fixed
committed, thanks!
> probably don't need to do full copy to row cache after un-mmap() change
> -----------------------------------------------------------------------
>
> Key: CASSANDRA-3223
> URL: https://issues.apache.org/jira/browse/CASSANDRA-3223
> Project: Cassandra
> Issue Type: Bug
> Components: Core
> Affects Versions: 1.0.0
> Reporter: Yang Yang
> Assignee: Yang Yang
> Priority: Minor
> Fix For: 1.0.0
>
> Attachments: 0002-do-not-need-to-deep-copy-column-value-bytebuffer-now.patch
>
>
> 3179 changes from directly using the bytebuffer from mmap(), to copying that buffer,
> CFS.cacheRow() https://github.com/apache/cassandra/blob/cassandra-1.0.0/src/java/org/apache/cassandra/db/ColumnFamilyStore.java line 1126
> says it makes a deep copy exactly to prevent issues from unmmap().
> maybe this deep copy is not needed now given 3179
> if so, maybe slightly better performance in both speed and memory
--
This message is automatically generated by JIRA.
For more information on JIRA, see: http://www.atlassian.com/software/jira