Thanks Bobby to enumerate the list. I agree we may want to sort out issues
in STORM-2714 <https://issues.apache.org/jira/browse/STORM-2714> (epic),
mostly linking real blockers to track how much we are close to 2.0.0.

Btw, adding to what Bobby lists up:

4) A JIRA to drop storm-kafka in favor of storm-kafka-client (STORM-2953
<https://issues.apache.org/jira/browse/STORM-2953>)

This is umbrella issue which Stig is working on. There're long pending pull
request to review: no one reviewed so far since Stig seems to be only
"active" PMC member who is an expert on storm-kafka-client. I may spend
time to review but as a non-expert, so it would be much better whoever
familiar with storm-kafka-client can review. For me this is a real blocker,
as we don't want to deal with storm-kafka anymore.

5) Review and fix/remove deprecated things in Storm 2.0.0 (STORM-2947
<https://issues.apache.org/jira/browse/STORM-2947>)

This is actually not a blocker but might be worth to address if we have
time to do it. If we haven't have time to do it, we can ignore the issue.
Not a big deal.

Happy to hear that there're nightly integration tests running against
master branch, since manual tests couldn't cover everything we need to
check. This will greatly reduce the stabilization period and then after
resolving blockers we are good to go with preparing release phase of 2.0.0.

Thanks,
Jungtaek Lim (HeartSaVioR)

2018년 6월 8일 (금) 오후 10:07, Bobby Evans <[email protected]>님이 작성:

> We have an Epic that is tracking the "blockers".
>
> https://issues.apache.org/jira/browse/STORM-2714
>
> But you have to be logged in to see any of the JIRA listed under it.  Some
> of them I don't consider to be true blockers.  Like moving the UI to java
> because we can do it afterwards without impacting users at all.  But I know
> that a pull request for it should show up shortly.
>
> For me personally the only real blockers left are not yet on the list.
>
> 1) I would like to go through an upgrade all of the dependencies for
> storm-client, and under external. STORM-3061
>
> 2) There is a netty 4 upgrade that has a pull request up STORM-1038
> <https://github.com/apache/storm/pull/2704>, but probably needs to be
> updated to deal with the shading change that just went in. (but if we do it
> right that could go in later as well because netty is shaded so it might
> not be a blocker either)
>
> 3) A JIRA to drop storm-druid in favor of using the kafka ingestion
> functionality in druid STORM-3097
> <https://github.com/apache/storm/pull/2707>.
>
> Some of STORM-3061 may be the long tail as it is going to need more
> testing time to be sure I didn't break any of them while upgrading their
> dependencies.
>
> So I would expect a release in a few weeks, partly because I am out on
> vacation next week so I wont be making a lot of progress on STORM-3097
> until I am back.  It is going to need some testing and bug fixes, but it is
> going to be a huge performance improvement for everyone.  The benchmark
> results are really looking good, but because everything is not done yet, I
> don't want to announce anything "official".
>
> Thanks,
>
> Bobby
>
>
> We have been running our integration tests on master nightly, but that is
> a compatibility mode where we are running older topologies on newer
> nimbus+supervisors and everything has been going more or less fine.  Some
> random failures that we are still tracking down and fixing as we go, but
> most of that is our tests having to deal with the code running faster than
> it did before.
>
> If we don't want to
>
>
>
> On Fri, Jun 8, 2018 at 3:37 AM Julien Nioche <
> [email protected]> wrote:
>
>> Hi,
>>
>> How close are we to having a 2.0.0 release?  Are there any blockers?
>>
>> Thanks
>>
>> Julien
>>
>> --
>>
>> *Open Source Solutions for Text Engineering*
>>
>> http://www.digitalpebble.com
>> http://digitalpebble.blogspot.com/
>> #digitalpebble <http://twitter.com/digitalpebble>
>>
>

Reply via email to