[jira] [Comment Edited] (CASSANDRA-6689) Partially Off Heap Memtables

["Benedict (JIRA)" <ji...@apache.org>]

    [ https://issues.apache.org/jira/browse/CASSANDRA-6689?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13913686#comment-13913686 ] 

Benedict edited comment on CASSANDRA-6689 at 2/26/14 11:08 PM:
---------------------------------------------------------------

bq. more granular to-the-point patches, plus it would be nice to merge this with 6694 so we can review/test all together

I don't want to seem obtuse here, I want to help as much as I can, but it really is one or the other. Even this patch would be difficult to break up, but 6694 makes it all even more intertwined. Note that excluding the minor suggestions made by [~krummas], 6694 is built on top of this, so contains this patch. I will be posting a patch shortly that addresses his concerns, but they're pretty small changes, so shouldn't affect the review process. I'll merge any changes with 6694 when I post.

If we can think of a good split to perform I'm all for it, but literally the only thing I can think of is to disable GC in the first commit. But if we commit 6694 (or this to only a slightly lesser extent) excluding GC, it's basically a waste of time to separate it, as it's a small fraction of the changes, and almost all of the complicated stuff needs to be done for disposal after flush without GC.


was (Author: benedict):
bq. more granular to-the-point patches, plus it would be nice to merge this with 6694 so we can review/test all together

I don't want to seem obtuse here, I want to help as much as I can, but it really is one or the other. Even this patch would be difficult to break up, but 6694 makes it all even more intertwined. Note that excluding the minor suggestions made by [~krummas], 6694 is built on top of this, so contains this patch. I will be posting a patch shortly that addresses his concerns, but they're pretty small changes, so shouldn't affect the review process. I'll merge any changes with 6694 when I post.

If we can think of a good split to perform I'm all for it, but literally the only thing I can think of is to disable GC in the first commit. But if we commit 6694 excluding GC, it's basically a waste of time to separate it, as it's a small fraction of the changes, and almost all of the complicated stuff needs to be done for disposal after flush without GC.

> Partially Off Heap Memtables
> ----------------------------
>
>                 Key: CASSANDRA-6689
>                 URL: https://issues.apache.org/jira/browse/CASSANDRA-6689
>             Project: Cassandra
>          Issue Type: New Feature
>          Components: Core
>            Reporter: Benedict
>            Assignee: Benedict
>             Fix For: 2.1 beta2
>
>
> Move the contents of ByteBuffers off-heap for records written to a memtable.
> (See comments for details)



--
This message was sent by Atlassian JIRA
(v6.1.5#6160)