[
https://issues.apache.org/jira/browse/CASSANDRA-2988?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13079088#comment-13079088
]
Stu Hood commented on CASSANDRA-2988:
-------------------------------------
bq. which totally applies to the index sampling both after a restart and when
opening a new streamed file.
Does it? If you are using and warming the keycache, then maybe you don't want
the index to be warm, but in any other case, the index should essentially be
locked in ram. Also, the streaming case is no longer linked to
SSTableReader.load in trunk.
bq. loading a new, streamed sstable (probably don't)
Same comment as above: .load() isn't involved.
----
Melvin: either way, you should post the numbers that you collected.
> Improve SSTableReader.load() when loading index files
> -----------------------------------------------------
>
> Key: CASSANDRA-2988
> URL: https://issues.apache.org/jira/browse/CASSANDRA-2988
> Project: Cassandra
> Issue Type: Improvement
> Reporter: Melvin Wang
>
> * when we create BufferredRandomAccessFile, we pass skipCache=true. This
> hurts the read performance because we always process the index files
> sequentially. Simple fix would be set it to false.
> * multiple index files of a single column family can be loaded in parallel.
> This buys a lot when you have multiple super large index files.
> * we may also change how we buffer. By using BufferredRandomAccessFile, for
> every read, we need bunch of checking like
> - do we need to rebuffer?
> - isEOF()?
> - assertions
> These can be simplified to some extent. We can blindly buffer the index
> file by chunks and process the buffer until a key lies across boundary of a
> chunk. Then we rebuffer and start from the beginning of the partially read
> key. Conceptually, this is same as what BRAF does but w/o the overhead in the
> read**() methods in BRAF.
--
This message is automatically generated by JIRA.
For more information on JIRA, see: http://www.atlassian.com/software/jira
