[jira] [Commented] (CASSANDRA-8630) Faster sequential IO (on compaction, streaming, etc)

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

    [ https://issues.apache.org/jira/browse/CASSANDRA-8630?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=14724615#comment-14724615 ] 

Stefania commented on CASSANDRA-8630:
-------------------------------------

Thanks [~benedict] and [~JoshuaMcKenzie].

bq. Removed the readBytes method entirely

Are you sure? This method was introduced by CASSANDRA-1714, titled _zero-copy reads_. It copies BBs via {{FastByteOperations}} rather than reading into a buffer byte by byte. Bear in mind that {{ByteBufferUtil.read()}} is one of the most used read methods according to flight recorder. 

bq. Removed the (as commented in TODO) unnecessary BufferedSegmentedFile.createReader

Thanks, I had totally forgot to remove this.

bq. MappedRegions.close now always calls channel.close, and returns any error thrown by it

Thanks for spotting it!

The remaining points are good. The new {{Throwables.perform}} API is really neat.

I've pushed a rebased version to Jenkins and I've asked on IRC to set-up Windows jobs for this branch. I've also launched a [cperf job|http://cstar.datastax.com/tests/id/35dbcc0e-5050-11e5-b3f6-42010af0688f]. I'll post the results once they are available.

> Faster sequential IO (on compaction, streaming, etc)
> ----------------------------------------------------
>
>                 Key: CASSANDRA-8630
>                 URL: https://issues.apache.org/jira/browse/CASSANDRA-8630
>             Project: Cassandra
>          Issue Type: Improvement
>          Components: Core, Tools
>            Reporter: Oleg Anastasyev
>            Assignee: Stefania
>              Labels: compaction, performance
>             Fix For: 3.x
>
>         Attachments: 8630-FasterSequencialReadsAndWrites.txt, cpu_load.png, flight_recorder_001_files.tar.gz, flight_recorder_002_files.tar.gz, mmaped_uncomp_hotspot.png
>
>
> When node is doing a lot of sequencial IO (streaming, compacting, etc) a lot of CPU is lost in calls to RAF's int read() and DataOutputStream's write(int).
> This is because default implementations of readShort,readLong, etc as well as their matching write* are implemented with numerous calls of byte by byte read and write. 
> This makes a lot of syscalls as well.
> A quick microbench shows than just reimplementation of these methods in either way gives 8x speed increase.
> A patch attached implements RandomAccessReader.read<Type> and SequencialWriter.write<Type> methods in more efficient way.
> I also eliminated some extra byte copies in CompositeType.split and ColumnNameHelper.maxComponents, which were on my profiler's hotspot method list during tests.
> A stress tests on my laptop show that this patch makes compaction 25-30% faster  on uncompressed sstables and 15% faster for compressed ones.
> A deployment to production shows much less CPU load for compaction. 
> (I attached a cpu load graph from one of our production, orange is niced CPU load - i.e. compaction; yellow is user - i.e. not compaction related tasks)



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