The regressions is already normalizing again. I'd observer it further before doing anything.

The same applies to the benchmarkCount which tanked even more in that same run.

On 15.03.2019 06:02, Tzu-Li (Gordon) Tai wrote:
@Yu
Thanks for reporting that Yu, great that this was noticed.

The serializerAvro case seems to only be testing on-wire serialization.
I checked the changes to the `AvroSerializer`, and it seems like
FLINK-11436 [1] with commit 479ebd59 was the only change that may have
affected that.
That commit wasn't introduced exactly around the time when the indicated
performance regression occurred, but was still before the regression.
The commit introduced some instanceof type checks / type casting in the
readObject of the AvroSerializer, which may have caused this.

Currently investigating further.

Cheers,
Gordon

On Fri, Mar 15, 2019 at 11:45 AM Yu Li <car...@gmail.com> wrote:

Hi Aljoscha and all,

 From our performance benchmark web site (
http://codespeed.dak8s.net:8000/changes/) I observed a noticeable
regression (-6.92%) on the serializerAvro case comparing the latest 100
revisions, which may need some attention. Thanks.

Best Regards,
Yu


On Thu, 14 Mar 2019 at 20:42, Aljoscha Krettek <aljos...@apache.org>
wrote:

Hi everyone,
Please review and vote on the release candidate 2 for Flink 1.8.0, as
follows:
[ ] +1, Approve the release
[ ] -1, Do not approve the release (please provide specific comments)


The complete staging area is available for your review, which includes:
* JIRA release notes [1],
* the official Apache source release and binary convenience releases to
be
deployed to dist.apache.org <http://dist.apache.org/> [2], which are
signed with the key with fingerprint
F2A67A8047499BBB3908D17AA8F4FD97121D7293 [3],
* all artifacts to be deployed to the Maven Central Repository [4],
* source code tag "release-1.8.0-rc2" [5],
* website pull request listing the new release [6]
* website pull request adding announcement blog post [7].

The vote will be open for at least 72 hours. It is adopted by majority
approval, with at least 3 PMC affirmative votes.

Thanks,
Aljoscha

[1]

https://issues.apache.org/jira/secure/ReleaseNote.jspa?projectId=12315522&version=12344274
<

https://issues.apache.org/jira/secure/ReleaseNote.jspa?projectId=12315522&version=12344274
[2] https://dist.apache.org/repos/dist/dev/flink/flink-1.8.0-rc2/ <
https://dist.apache.org/repos/dist/dev/flink/flink-1.8.0-rc2/>
[3] https://dist.apache.org/repos/dist/release/flink/KEYS <
https://dist.apache.org/repos/dist/release/flink/KEYS>
[4]
https://repository.apache.org/content/repositories/orgapacheflink-1213
<https://repository.apache.org/content/repositories/orgapacheflink-1210/

[5]

https://gitbox.apache.org/repos/asf?p=flink.git;a=tag;h=c77a329b71e3068bfde965ae91921ad5c47246dd
<

https://gitbox.apache.org/repos/asf?p=flink.git;a=tag;h=2d00b1c26d7b4554707063ab0d1d6cc236cfe8a5
[6] https://github.com/apache/flink-web/pull/180 <
https://github.com/apache/flink-web/pull/180>
[7] https://github.com/apache/flink-web/pull/179 <
https://github.com/apache/flink-web/pull/179>

P.S. The difference to the previous RC1 is very small, you can fetch the
two tags and do a "git log release-1.8.0-rc1..release-1.8.0-rc2” to see
the
difference in commits. Its fixes for the issues that led to the
cancellation of the previous RC plus smaller fixes. Most
verification/testing that was carried out should apply as is to this RC.


Reply via email to