[jira] [Commented] (CASSANDRA-13241) Lower default chunk_length_in_kb from 64kb to 4kb

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

    [ https://issues.apache.org/jira/browse/CASSANDRA-13241?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=15889653#comment-15889653 ] 

Benjamin Roth commented on CASSANDRA-13241:
-------------------------------------------

I thought of 2 arrays because a semantic meaning (position vs chunk size) and a single alignment (8, 3, 2 byte) for each could be easier to understand and to maintain. Of course it works either way. With 2 arrays, you could still "pull sections", it's just a single fetch more to get the 8 byte absolute offset.
Loop summing vs. "relative-absolute offset": At the end this is always a tradeoff between mem/cpu. I personally am not the one who fights for every single byte in this case. But I also think some CPU cycles more to sum a bunch of ints is still bearable. I guess if I had to decide, I'd give "loop summing" a try. Any different opinions?

Do you mean a ChunkCache cache miss? Sorry for that kind of questions. I never came across this part of the code.

> Lower default chunk_length_in_kb from 64kb to 4kb
> -------------------------------------------------
>
>                 Key: CASSANDRA-13241
>                 URL: https://issues.apache.org/jira/browse/CASSANDRA-13241
>             Project: Cassandra
>          Issue Type: Wish
>          Components: Core
>            Reporter: Benjamin Roth
>
> Having a too low chunk size may result in some wasted disk space. A too high chunk size may lead to massive overreads and may have a critical impact on overall system performance.
> In my case, the default chunk size lead to peak read IOs of up to 1GB/s and avg reads of 200MB/s. After lowering chunksize (of course aligned with read ahead), the avg read IO went below 20 MB/s, rather 10-15MB/s.
> The risk of (physical) overreads is increasing with lower (page cache size) / (total data size) ratio.
> High chunk sizes are mostly appropriate for bigger payloads pre request but if the model consists rather of small rows or small resultsets, the read overhead with 64kb chunk size is insanely high. This applies for example for (small) skinny rows.
> Please also see here:
> https://groups.google.com/forum/#!topic/scylladb-dev/j_qXSP-6-gY
> To give you some insights what a difference it can make (460GB data, 128GB RAM):
> - Latency of a quite large CF: https://cl.ly/1r3e0W0S393L
> - Disk throughput: https://cl.ly/2a0Z250S1M3c
> - This shows, that the request distribution remained the same, so no "dynamic snitch magic": https://cl.ly/3E0t1T1z2c0J



--
This message was sent by Atlassian JIRA
(v6.3.15#6346)