activemq-dev mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From Kevin Burton <bur...@spinn3r.com>
Subject Re: ActiveMQ 6.0.0 required Java 8?
Date Thu, 19 Mar 2015 02:18:26 GMT
Yes.  Sort of.  There was a regression for persistent=false which breaks it
for advisories.

https://issues.apache.org/jira/browse/AMQ-5665

I was *hoping* it fixed the issue.

If it doesn’t I was going to write a test and then git bisect to find where
it broke.

One other problem I ran into:

https://github.com/apache/activemq

Does not have the source for 6.0.0 (unless I’m missing something.)

master is 5.11 snapshot and there are no 6.0.0 branches..

The other issue I had, was that a lot of the modules changed.  So I was
trying to track down the source to figure out which modules have been
renamed but of course I can’t find the source :-P

We’re still trying to deploy a pretty large ActiveMQ install.  Right now
it’s on 8 servers and has about 80GB of messages.  5.10.x has had a number
of issues for us.  I fixed two significant ones but they weren’t merged for
6.0.0.  The pull request was for 5.10.x and 5.11.x but it seems to have
been left behind? It was about 2 days worth of work and fixes a pretty
major scalability issue for ActiveMQ with a large number of queues.

I’m also pretty convinced I’ve found another bug whereby the entire queue
serves messages at about 1/100th the correct speed and queues grow very
large with nothing being served.  I was going to try to get on 5.11 or
6.0.0 but I can’t with the above bug in advisories.

I don’t mind stepping in and fixing these issues btw.  But I need to figure
out the right way to contribute so my pull requests don’t go into
purgatory.  Not pointing figures.. I just need to figure out a way to avoid
having my work left behind.

Maybe officially rejecting the pull request with a reason would help?

Purgatory and lost work seems to be a far worst situation than a ‘no, we’re
not going to merge that because of X’ because I can fix this situation! :)

If I know how to resolve these I’ll take my patches out of the graveyard
and port them to 6.0.0 and then get the AMQ-5665 fixed and get a pull
request for that as well.

Kevin




On Wed, Mar 18, 2015 at 7:02 PM, Clebert Suconic <clebert.suconic@gmail.com>
wrote:

> At this point we need java8 to build the source, but the target
> compilation still java 1.7
>
> We are not using any java8 features at this point. We kind of stepped
> back on being strict about java8. We could have updated the docs but
> since we kept java8 to build the source we are still recommending
> java8.
>
>
> The testsuite is running on java8 now, but it has been on java7 up
> till recently.
>
>
> I would recommend java8 as java7 is almost EOL but it still safe to
> use java7 on the binaries at this point.
>
>
> Are you evaluating it already? We are looking for feedback about it...
> we are still under voting for the release.. so any feedback helps!
>
>
>
>
> On Wed, Mar 18, 2015 at 9:48 PM, Kevin Burton <burton@spinn3r.com> wrote:
> > Just curious.  We’re still on Java 1.7.  I assume Java 8 features are
> > actually used.  Might be bad news for us but I can see it being a
> > reasonable requirement at this point.
> >
> > --
> >
> > Founder/CEO Spinn3r.com
> > Location: *San Francisco, CA*
> > blog: http://burtonator.wordpress.com
> > … or check out my Google+ profile
> > <https://plus.google.com/102718274791889610666/posts>
> > <http://spinn3r.com>
>
>
>
> --
> Clebert Suconic
> http://community.jboss.org/people/clebert.suconic@jboss.com
> http://clebertsuconic.blogspot.com
>



-- 

Founder/CEO Spinn3r.com
Location: *San Francisco, CA*
blog: http://burtonator.wordpress.com
… or check out my Google+ profile
<https://plus.google.com/102718274791889610666/posts>
<http://spinn3r.com>

Mime
  • Unnamed multipart/alternative (inline, None, 0 bytes)
View raw message