Thanks Ariel & Jonathan.
Michael, I have addressed the comments in the issue so we're using the latest
jar now. I've posted the updated patch in the latest comment of the ticket. In
addition there is a JMH performance benchmark in the C* repo
(CompressorPerformance) that tests various Compressors. Depending on the
`compression_level` setting we can see a boost in compression speed. However,
the major win with Zstd is not only the speed but the superior compression
ratio. While lz4 can achieve about 2.101, Zstd can achieve a ratio of 2.877
which is a huge difference.
The comparisons are here - https://github.com/facebook/zstd
Dinesh
On Friday, February 15, 2019, 9:56:49 AM PST, Michael Shuler
<[email protected]> wrote:
+0.5
I skimmed the jira and github diff and a few things came to mind:
- There are multiple comments about using an older jar than the latest
version.
- I did not see any performance test results to form an opinion on any
gains/caveats as a user. This was the first thing I looked for.
- I did not see any conf/cassandra.yaml comment in the diff for the
valid class_name configuration option to use.
Seems interesting, and there are comments about 4.0 being in a freeze
and all, but OK on it being non-default.
--
Michael
On 2/15/19 11:43 AM, Ariel Weisberg wrote:
> Hi,
>
> I am +1 since it's an additional compressor and not the default.
>
> Ariel
>
> On Fri, Feb 15, 2019, at 11:41 AM, Dinesh Joshi wrote:
>> Hey folks,
>>
>> Just wanted to get a pulse on whether we can proceed with ZStd support.
>> The consensus on the ticket was that it’s a very valuable addition
>> without any risk of destabilizing 4.0. It’s ready to go if there aren’t
>> any objections.
>>
>> Dinesh
>>
>> ---------------------------------------------------------------------
>> To unsubscribe, e-mail: [email protected]
>> For additional commands, e-mail: [email protected]
>>
>
> ---------------------------------------------------------------------
> To unsubscribe, e-mail: [email protected]
> For additional commands, e-mail: [email protected]
>
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
