[
https://issues.apache.org/jira/browse/CASSANDRA-8894?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=14600585#comment-14600585
]
Stefania commented on CASSANDRA-8894:
-------------------------------------
I pushed a commit to fix the rounding, thanks.
I've run a couple of cperf tests:
- [one|http://cstar.datastax.com/tests/id/27ca75a2-1ae0-11e5-bede-42010af0688f]
comparing trunk and 8894 both with disk_access_mode standard, here we can see
8894 is better when reading (~6000 ops per second).
- [one|http://cstar.datastax.com/tests/id/e1d62922-1ae1-11e5-9038-42010af0688f]
comparing trunk mmap and 8894 standard, here we can see trunk is still better
than 8894 when reading (~4000 ops per second).
CI is still running.
> Our default buffer size for (uncompressed) buffered reads should be smaller,
> and based on the expected record size
> ------------------------------------------------------------------------------------------------------------------
>
> Key: CASSANDRA-8894
> URL: https://issues.apache.org/jira/browse/CASSANDRA-8894
> Project: Cassandra
> Issue Type: Improvement
> Components: Core
> Reporter: Benedict
> Assignee: Stefania
> Fix For: 3.x
>
>
> A large contributor to slower buffered reads than mmapped is likely that we
> read a full 64Kb at once, when average record sizes may be as low as 140
> bytes on our stress tests. The TLB has only 128 entries on a modern core, and
> each read will touch 32 of these, meaning we are unlikely to almost ever be
> hitting the TLB, and will be incurring at least 30 unnecessary misses each
> time (as well as the other costs of larger than necessary accesses). When
> working with an SSD there is little to no benefit reading more than 4Kb at
> once, and in either case reading more data than we need is wasteful. So, I
> propose selecting a buffer size that is the next larger power of 2 than our
> average record size (with a minimum of 4Kb), so that we expect to read in one
> operation. I also propose that we create a pool of these buffers up-front,
> and that we ensure they are all exactly aligned to a virtual page, so that
> the source and target operations each touch exactly one virtual page per 4Kb
> of expected record size.
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
