[ 
https://issues.apache.org/jira/browse/CASSANDRA-9738?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=14702731#comment-14702731
 ] 

Robert Stupp commented on CASSANDRA-9738:
-----------------------------------------

Thanks :)

bq. How are the 2i paths tested?

Currently implicitly via other unit tests. But I'll add some unit test for that.

bq. The null case in makeVal isn't tested, maybe not that interesting

Oh. Although it doesn't cause any problems, passing null to OHC is not allowed. 
I've added an assertion.

bq. SerializationHeader forKeyCache

Yes. It's intentionally racy to prevent blocking operations. The chance for 
such a race is probably low and "heals" itself. Added a comment for that.

bq. comment about singleton weigher

removed

bq. NIODataInputStream has a derived class DataInputBuffer

changed to use DIB

bq. string encoding and decoding helpers

yup - makes sense. refactored.

bq. An enhancement we can file for later is to replace those strings with vints 
that reference a map of possible table names.

My idea is to remove strings at all from the key cache. Keyspace + CF names can 
be handled by CASSANDRA-10028. Not sure how to handle file/path names - maybe 
using some sparse list structure for sstable generations (in another ticket).

Haven't pushed anything yet - but will update my branch soon.

> Migrate key-cache to be fully off-heap
> --------------------------------------
>
>                 Key: CASSANDRA-9738
>                 URL: https://issues.apache.org/jira/browse/CASSANDRA-9738
>             Project: Cassandra
>          Issue Type: Sub-task
>            Reporter: Robert Stupp
>            Assignee: Robert Stupp
>             Fix For: 3.0 beta 2
>
>
> Key cache still uses a concurrent map on-heap. This could go to off-heap and 
> feels doable now after CASSANDRA-8099.
> Evaluation should be done in advance based on a POC to prove that pure 
> off-heap counter cache buys a performance and/or gc-pressure improvement.
> In theory, elimination of on-heap management of the map should buy us some 
> benefit.



--
This message was sent by Atlassian JIRA
(v6.3.4#6332)

Reply via email to