[Announce] Gerrit 2.5-rc2 (release candidate)

[Edwin Kempin]

Hi,

I am pleased to announce Gerrit Code Review 2.5-rc2

Download:

  http://code.google.com/p/gerrit/downloads/detail?name=gerrit-full-2.5-rc2.war

This is the third release candidate for Gerrit 2.5.

Release Notes:
  http://gerrit-documentation.googlecode.com/svn/ReleaseNotes/ReleaseNotes-2.5.html

If there are no major issues with this release candidate
it should become the final Gerrit 2.5 release.

I plan to release the final Gerrit 2.5 version in November
before the Gerrit User Summit.

Here is a short summary of what has changed between Gerrit-2.5-rc1
and Gerrit-2.5-rc2.

Bug-Fixes:
* Fix nested submodule updates
* Fix that 'draft-published' event is not visible
* Fix 'test-submit-rule' SSH command
* Fix NPE in 'query' SSH command when used with the options '--comments --format=JSON'
* Ensure that RevWalk in SubmoduleOp is released

Performance:
* Add LDAP-cache to minimize number of queries when unnesting groups

Documentation Updates:
* Update release notes for Gerrit 2.5-rc2
* Warn in release notes that internal groups can't contain external groups
* Explain warning for schema version 68 in Gerrit 2.5 release notes

Edwin

[Manuel Doninger]

Hi,
is there any chance to get one of the proposed fixes for the issue
mentioned in https://groups.google.com/forum/?fromgroups=#!topic/repo-discuss/OP6hg1KYah8/discussion
into the release?
We want to use HTTP access to avoid the administrative overhead with
maintaining SSH Keys and access, and are facing that issue.

Regards,
Manuel

> --
> To unsubscribe, email repo-discuss...@googlegroups.com
> More info at http://groups.google.com/group/repo-discuss?hl=en

[Martin Fick]

On Thursday, October 25, 2012 09:13:41 am Manuel Doninger
wrote:

> Hi,
> is there any chance to get one of the proposed fixes for
> the issue mentioned in
> https://groups.google.com/forum/?fromgroups=#!topic/repo
> -discuss/OP6hg1KYah8/discussion into the release?
> We want to use HTTP access to avoid the administrative
> overhead with maintaining SSH Keys and access, and are
> facing that issue.

I don't believe that any solutions have been posted for
review (uploaded as changes) for this, and it is getting
late in the rc cycle. I would advise against holding off
for this since it does not appear to be a new bug.

I would like to propose that we finalize 2.5 as soon as
possible and then release 2.6-rc0 right before the hackathon
since there are already over 200 changes on master which do
not appear to be in 2.5-rc2. This way there will be no need
to wait for all the hackathon code to stabilize before
releasing 2.6,

-Martin

[Shawn Pearce]

Unfortunately... I agree with Martin. :-\

[Edwin Kempin]

2012/10/25 Shawn Pearce <s...@google.com>

On Thu, Oct 25, 2012 at 8:56 AM, Martin Fick <mf...@codeaurora.org> wrote:
> On Thursday, October 25, 2012 09:13:41 am Manuel Doninger
> wrote:
>> Hi,
>> is there any chance to get one of the proposed fixes for
>> the issue mentioned in
>> https://groups.google.com/forum/?fromgroups=#!topic/repo
>> -discuss/OP6hg1KYah8/discussion into the release?
>> We want to use HTTP access to avoid the administrative
>> overhead with maintaining SSH Keys and access, and are
>> facing that issue.
>
> I don't believe that any solutions have been posted for
> review (uploaded as changes) for this, and it is getting
> late in the rc cycle.  I would advise against holding off
> for this since it does not appear to be a new bug.
>
> I would like to propose that we finalize 2.5 as soon as
> possible and then release 2.6-rc0 right before the hackathon

+1
 

> since there are already over 200 changes on master which do
> not appear to be in 2.5-rc2.  This way there will be no need
> to wait for all the hackathon code to stabilize before
> releasing 2.6,

Unfortunately... I agree with Martin. :-\

[Edwin Kempin]

2012/10/25 Edwin Kempin <edwin....@gmail.com>

I'm about to do the final 2.5 release, but it feels bad to me that push via
HTTP is broken. I agree that it's probably to late to do a big analysis and
bug-fixing, but I would like to propose another solution. Commit
c545c0901241314190cac02a24aa95f831dd0572 [1] which seems to have
broken push over HTTP seems to be relevant for Shawn only. Reverting
this commit fixes the problem. Since Shawn is anyway not running 2.5
but master, I think it should be ok to revert this commit in stable-2.5 [2], but
keeping it in master. Until 2.6 we have then enough time to come up with a
fix in master. What do you think?

[1] https://gerrit-review.googlesource.com/37061
[2] https://gerrit-review.googlesource.com/39090

For the record, here is the stacktrace:
[2012-10-31 13:56:21,457] ERROR com.google.gerrit.server.git.ReceiveCommits : Can't insert changes for myProject
com.google.gwtorm.server.OrmException: Error updating database
    at com.google.gerrit.server.git.ReceiveCommits$1.apply(ReceiveCommits.java:226)
    at com.google.gerrit.server.git.ReceiveCommits$1.apply(ReceiveCommits.java:220)
    at com.google.common.util.concurrent.Futures$MappingCheckedFuture.mapException(Futures.java:1215)
    at com.google.common.util.concurrent.AbstractCheckedFuture.checkedGet(AbstractCheckedFuture.java:85)
    at com.google.gerrit.server.git.ReceiveCommits.insertChangesAndPatchSets(ReceiveCommits.java:675)
    at com.google.gerrit.server.git.ReceiveCommits.processCommands(ReceiveCommits.java:521)
    at com.google.gerrit.server.git.AsyncReceiveCommits$Worker.run(AsyncReceiveCommits.java:90)
    at java.util.concurrent.Executors$RunnableAdapter.call(Unknown Source)
    at com.google.gerrit.server.util.RequestScopePropagator$5.call(RequestScopePropagator.java:196)
    at com.google.gerrit.server.util.RequestScopePropagator$4.call(RequestScopePropagator.java:174)
    at com.google.inject.servlet.ServletScopes$3.call(ServletScopes.java:194)
    at com.google.gerrit.server.util.RequestScopePropagator$1.call(RequestScopePropagator.java:81)
    at com.google.gerrit.server.util.RequestScopePropagator$2.run(RequestScopePropagator.java:113)
    at java.util.concurrent.Executors$RunnableAdapter.call(Unknown Source)
    at java.util.concurrent.FutureTask$Sync.innerRun(Unknown Source)
    at java.util.concurrent.FutureTask.run(Unknown Source)
    at java.util.concurrent.ScheduledThreadPoolExecutor$ScheduledFutureTask.access$301(Unknown Source)
    at java.util.concurrent.ScheduledThreadPoolExecutor$ScheduledFutureTask.run(Unknown Source)
    at com.google.gerrit.server.git.WorkQueue$Task.run(WorkQueue.java:337)
    at java.util.concurrent.ThreadPoolExecutor$Worker.runTask(Unknown Source)
    at java.util.concurrent.ThreadPoolExecutor$Worker.run(Unknown Source)
    at java.lang.Thread.run(Unknown Source)
Caused by: java.util.concurrent.ExecutionException: java.lang.IllegalStateException: Cannot continue request in the same thread as a HTTP request!
    at java.util.concurrent.FutureTask$Sync.innerGet(Unknown Source)
    at java.util.concurrent.FutureTask.get(Unknown Source)
    at com.google.common.util.concurrent.ForwardingFuture.get(ForwardingFuture.java:63)
    at com.google.common.util.concurrent.AbstractCheckedFuture.checkedGet(AbstractCheckedFuture.java:78)
    ... 18 more
Caused by: java.lang.IllegalStateException: Cannot continue request in the same thread as a HTTP request!
    at com.google.inject.internal.util.$Preconditions.checkState(Preconditions.java:142)
    at com.google.inject.servlet.ServletScopes$3.call(ServletScopes.java:187)
    at com.google.gerrit.server.util.RequestScopePropagator$1.call(RequestScopePropagator.java:81)
    at java.util.concurrent.FutureTask$Sync.innerRun(Unknown Source)
    at java.util.concurrent.FutureTask.run(Unknown Source)
    at com.google.common.util.concurrent.MoreExecutors$SameThreadExecutorService.execute(MoreExecutors.java:253)
    at com.google.common.util.concurrent.AbstractListeningExecutorService.submit(AbstractListeningExecutorService.java:56)
    at com.google.gerrit.server.git.ReceiveCommits$CreateRequest.insertChange(ReceiveCommits.java:1291)
    at com.google.gerrit.server.git.ReceiveCommits.insertChangesAndPatchSets(ReceiveCommits.java:671)
    ... 17 more


 

[Shawn Pearce]

On Wed, Oct 31, 2012 at 6:16 AM, Edwin Kempin <edwin....@gmail.com> wrote:
> I'm about to do the final 2.5 release, but it feels bad to me that push via
> HTTP is broken. I agree that it's probably to late to do a big analysis and
> bug-fixing, but I would like to propose another solution. Commit
> c545c0901241314190cac02a24aa95f831dd0572 [1] which seems to have
> broken push over HTTP seems to be relevant for Shawn only. Reverting
> this commit fixes the problem. Since Shawn is anyway not running 2.5
> but master, I think it should be ok to revert this commit in stable-2.5 [2],
> but
> keeping it in master. Until 2.6 we have then enough time to come up with a
> fix in master. What do you think?
>
> [1] https://gerrit-review.googlesource.com/37061
> [2] https://gerrit-review.googlesource.com/39090

I just +2'd the revert in stable-2.5.

Please be careful merging stable-2.5 into master. We don't want the
revert merging in. I would merge the parent of this commit to master,
and then -s ours merge the commit to master.

[Edwin Kempin]

2012/10/31 Shawn Pearce <s...@google.com>

Thanks, I've done the merges as you said:
  https://gerrit-review.googlesource.com/39110
  https://gerrit-review.googlesource.com/39111