[
https://issues.apache.org/jira/browse/CASSANDRA-16360?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=17335219#comment-17335219
]
Alexey Zotov commented on CASSANDRA-16360:
------------------------------------------
Oh ok, I got your point. You've been talking about lookup table caching rather
than data caching. That makes perfect sense to me now. I definitely agree with
your points regarding complexity of its testing and lack of the real necessity
to do it. Therefore:
* I'm parking the benchmark code as is (unless some changes are requested)
* I'm waiting for the decision whether I can currently proceed with step 7
from [~samt]
* I'll be waiting for 4.x where we have no Java 8 support
PS:
{quote}That said, you may well be benchmarking on a machine without crc32c
hardware support, anyway. Which would explain the very similar performance
profile.
{quote}
That's what I've checked on the early stage of the local benchmarking. I have
[i7-7500u|https://ark.intel.com/content/www/us/en/ark/products/95451/intel-core-i7-7500u-processor-4m-cache-up-to-3-50-ghz.html]
which support SSE4.2 instructions set which in turn contains [CRC
operation|https://en.wikipedia.org/wiki/SSE4#SSE4.2] for _CRC32C_ calculation.
> CRC32 is inefficient on x86
> ---------------------------
>
> Key: CASSANDRA-16360
> URL: https://issues.apache.org/jira/browse/CASSANDRA-16360
> Project: Cassandra
> Issue Type: Improvement
> Components: Messaging/Client
> Reporter: Avi Kivity
> Assignee: Alexey Zotov
> Priority: Normal
> Labels: protocolv6
> Fix For: 4.0.x
>
>
> The client/server protocol specifies CRC24 and CRC32 as the checksum
> algorithm (cql_protocol_V5_framing.asc). Those however are expensive to
> compute; this affects both the client and the server.
>
> A better checksum algorithm is CRC32C, which has hardware support on x86 (as
> well as other modern architectures).
--
This message was sent by Atlassian Jira
(v8.3.4#803005)
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
