[
https://issues.apache.org/jira/browse/CASSANDRA-6689?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13923770#comment-13923770
]
Benedict commented on CASSANDRA-6689:
-------------------------------------
bq. One native Buffer each for name and value.
Probably we'll stick to a Buffer per name component here, rather than per name,
for simplicity at this step.
This timeline translates to mostly what I suggested already, with a little
tweaking, so I'll work with that (I have a new step 1, though):
# I will introduce the -> .data refactor as my starting point, and refactors of
.memory in 6694, all ignoring any Native implementations;
# Introduce a NativeByteBufferAllocator, and always copy onto heap if reading
from it
# Introduce NativeCell, and a simplified NativeAllocator (i.e. without any
magic to reclaim it)
# Introduce RefAction + RefAction driven reclaim
# Introduce 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
>
> Attachments: CASSANDRA-6689-small-changes.patch
>
>
> Move the contents of ByteBuffers off-heap for records written to a memtable.
> (See comments for details)
--
This message was sent by Atlassian JIRA
(v6.2#6252)
