RE: Backports for LTS 1.424.3

[Nord, James]

Unfortunately this doesn't fix it (sorry only just been able to test)

Jenkins 1.424.6

INFO: Started initialization
Mar 08, 2012 1:47:27 PM jenkins.model.Jenkins$6 onAttained
INFO: Listed all plugins
log4j:WARN No appenders could be found for logger (net.bull.javamelody).
log4j:WARN Please initialize the log4j system properly.
SLF4J: Failed to load class "org.slf4j.impl.StaticLoggerBinder".
SLF4J: Defaulting to no-operation (NOP) logger implementation
SLF4J: See http://www.slf4j.org/codes.html#StaticLoggerBinder for further 
details.
Mar 08, 2012 1:47:30 PM jenkins.model.Jenkins$6 onAttained
INFO: Prepared all plugins

:-(

I'm taking a guess that the maven plugin is loaded first and the slf4j 
implementation is provided by the Gerrit and m2release plugins.  But the 
classloaders haven't been added so when slf4j is loaded for maven job type it 
can't find an implementation and fails.  So latter when the Gerrit/m2release 
plugin load they get the same static do nothing logger.

[root@jenkins_inst plugins]# find . -name \*slf\*.jar
./m2release/WEB-INF/lib/slf4j-jdk14-1.6.0.jar
./m2release/WEB-INF/lib/slf4j-api-1.6.0.jar
./maven-plugin/WEB-INF/lib/slf4j-api-1.6.1.jar
./gerrit-trigger/WEB-INF/lib/slf4j-jdk14-1.6.1.jar
./gerrit-trigger/WEB-INF/lib/slf4j-api-1.6.1.jar

So the binding would need to be present by default in the main webapp 
WEB-INF/lib/ - or provided by the plugin.

I'm aware there is an issue in JIRA about problems with providing the binding 
with a certain container - but really no logs by default out of the box has to 
be worse than blowing up because someone can't write a container correctly 
doesn't it?

/James


> -Original Message-
> From: jenkinsci-dev@googlegroups.com [mailto:jenkinsci-
> d...@googlegroups.com] On Behalf Of Vojtech Juranek
> Sent: 02 February 2012 20:37
> To: jenkinsci-dev@googlegroups.com
> Subject: Re: Backports for LTS 1.424.3
>
> Hi James,
>
> > An easier test is to open up the maven-plugin.hpi If it contains
> > \WEB-INF\lib\slf4j-noop-*.jar then it will prevent logging
>
> thanks for easy test, if it's just about presence of slf4j-nop-*.jar (I guess 
> you
> meant slf4j-not, not noop), it can be fixed by upgrade to wagon 2.2 - it's 
> this
> series of commits:
> 532ab9448de62920fd5e2dcb73c2e0b024f39d11
> ec29ce6f0c8e3a77aae54c485a28c741a266dc93
> 4f044c282bb2473c41a4fca32dbbd1f46d276995
> 7bba3abcf9456e5a933e2207c4d87bc58a9f2fc0
>
> hopefully it doesn't break anything else...
>
> > But a trawl of JIRA shows that it's probably not even been fixed (and
> > has even got worse!
>
> looks like it was introduced in sshd, which is on in 1.424
>
> > - except for some questionable container like EAP that doesn't isolate
> > classloaders correctly!)
>
> :-D


**
This message is confidential and intended only for the addressee. If you have 
received this message in error, please immediately notify the 
postmas...@nds.com and delete it from your system as well as any copies. The 
content of e-mails as well as traffic data may be monitored by NDS for 
employment and security purposes. To protect the environment please do not 
print this e-mail unless necessary.

NDS Limited. Registered Office: One London Road, Staines, Middlesex, TW18 4EX, 
United Kingdom. A company registered in England and Wales. Registered no. 
3080780. VAT no. GB 603 8808 40-00
**

Re: Test status of 1.424.3 RC2 [WAS] Re: Backports for LTS 1.424.3

[Vojtech Juranek]

> Sorry for being late in the party, but the update center URL is kept
> the same between LTS and main line intentionally.
 
> The server does the redirect based on the version of Jenkins that's
> requesting the update center data.  This is necessary to seamlessly
> handle the situation where people move between mainline and LTS.

now make sense to me (there hasn't been any "/stable" URL before and I was 
thinking, without any success, how it actually works yesterday:-), I've 
reverted the  URL back to original value

> The real issue here is that when you restart Jenkins by switching from
> mainline to LTS (or vice versa), we have no means to refresh the new
> update center metadata right away, resulting in the phenomena that
> Ulli observed until the next scheduled update happens.
> 
> Maybe we can force the retrieval of the metadata upon the boot (if the
> version number is a released version, so as not to cause unnecessary
> retrieval during development.)
> 
> So my vote is to stick with RC3 and fix this in the next release.

as it's not as trivial as change of one URL, completely agree (btw: you 
probably meant RC2 from Feb. 14th, didn't you)

Re: Test status of 1.424.3 RC2 [WAS] Re: Backports for LTS 1.424.3

[Kohsuke Kawaguchi]

Sorry for being late in the party, but the update center URL is kept
the same between LTS and main line intentionally.

The server does the redirect based on the version of Jenkins that's
requesting the update center data. This is necessary to seamlessly
handle the situation where people move between mainline and LTS.

The real issue here is that when you restart Jenkins by switching from
mainline to LTS (or vice versa), we have no means to refresh the new
update center metadata right away, resulting in the phenomena that
Ulli observed until the next scheduled update happens.

Maybe we can force the retrieval of the metadata upon the boot (if the
version number is a released version, so as not to cause unnecessary
retrieval during development.)

So my vote is to stick with RC3 and fix this in the next release.

2012/2/19 Vojtech Juranek :
> Hi,
>
>> Thanks for catching this. Respinning.
>
> I re-run most of the tests and looks good (including java auto-install), i.e.
> IMHO ready for release.
>
> Ulli spotted one more bug. As I mentioned early today [1], we should include
> bug fix without re-testing whole release or postpone bug fix till next 
> release,
> not to delay the release.
> Personally, I would choose first option.
>
> Cheers
> Vojta
>
> [1] http://groups.google.com/group/jenkinsci-dev/msg/55ea1e04e99a0c64



-- 
Kohsuke Kawaguchi

Re: Test status of 1.424.3 RC2 [WAS] Re: Backports for LTS 1.424.3

[Ulli Hafner]

On 02/19/2012 02:22 PM, Vojtech Juranek wrote:

Hi,


Thanks for catching this. Respinning.

I re-run most of the tests and looks good (including java auto-install), i.e.
IMHO ready for release.

Ulli spotted one more bug. As I mentioned early today [1], we should include
bug fix without re-testing whole release or postpone bug fix till next release,
not to delay the release.
Personally, I would choose first option.

Cheers
Vojta

[1] http://groups.google.com/group/jenkinsci-dev/msg/55ea1e04e99a0c64
I also think that option [1] is best, since the required change is just 
a simple URL change...


Ulli

Test status of 1.424.3 RC2 [WAS] Re: Backports for LTS 1.424.3

[Vojtech Juranek]

Hi,

> Thanks for catching this. Respinning.

I re-run most of the tests and looks good (including java auto-install), i.e. 
IMHO ready for release.

Ulli spotted one more bug. As I mentioned early today [1], we should include 
bug fix without re-testing whole release or postpone bug fix till next release, 
not to delay the release.
Personally, I would choose first option.

Cheers
Vojta

[1] http://groups.google.com/group/jenkinsci-dev/msg/55ea1e04e99a0c64

Re: Backports for LTS 1.424.3

[Vojtech Juranek]

Hi,

> Shouldn't the LTS version only have upgrades available to newer LTS
> versions? (At least the message should note that this is not an LTS
> release)

thanks for spotting it, fixed [1]. As I'll finish testing of previous re-spin 
during today and LTS release is already delayed, I would propose:

* include this fix without testing (trivial change, low probability, that it 
breaks anything, automated test were passed [2])
OR
* skip it and include it in next release, as it's not a blocker and can be 
easily fixed manually in case of need (change 
$JENKINS_HOME/hudson.model.UpdateCenter.xml)
 
> Also the bundled plug-ins have quite old versions, are these later
> versions that are also marked stable? (e.g. CVS 1.6 instead of 1.2).
> When I upgrade these plug-ins and restart Jenkins then (sometimes) the
> old version is still present. Is this a bug or intended?

I think in LTS we upgrade anything only when needed (i.e. because of serious 
bug fix requires it).
As for re-writing by old version, AFAIK this is not specific to LTS and not 
sure if it's bug or intentional (my understanding is that's intentional)

Cheers
Vojta

[1] 
https://github.com/jenkinsci/jenkins/commit/6844b22c5c93df80694f6e8d1aa29777874e951c
[2] http://ci.jenkins-ci.org/view/Jenkins%20core/job/jenkins_lts_branch/44/

Re: Backports for LTS 1.424.3

[Ullrich Hafner]

When I'm running the RC then the message 'New version of Jenkins (1.451)
is available for download

(changelog ).' appears on the
manage screen.

Shouldn't the LTS version only have upgrades available to newer LTS
versions? (At least the message should note that this is not an LTS release)

Also the bundled plug-ins have quite old versions, are these later
versions that are also marked stable? (e.g. CVS 1.6 instead of 1.2).
When I upgrade these plug-ins and restart Jenkins then (sometimes) the
old version is still present. Is this a bug or intended?

Ulli


On 02/14/2012 07:31 PM, Kohsuke Kawaguchi wrote:
>
> Thanks for catching this. Respinning.
>
> On 02/13/2012 01:56 PM, Vojtech Juranek wrote:
>> Hi,
>> during the testing I found out that auto-install of JDK doesn't work
>> in most
>> of the cases (it works for jdk 1.7.0 and as I tested it in against
>> 1.7.0 I
>> didn't spot it  in previous releases). I track down that it was
>> logged as
>> JENKINS-11420 and fixed  quite long time ago.
>>
>> It's marked as "major", but I would consider it as quite important
>> feature and
>> therefore I backported the fix. Could you please re-spin? It passed all
>> automated tests [1].
>>
>> Thanks
>> Vojta
>>
>> [1]
>> http://ci.jenkins-ci.org/view/Jenkins%20core/job/jenkins_lts_branch/43/
>>
>>>  Done. It's at their usual places: http://mirrors.jenkins-ci.org/
>>
>
>

Re: Backports for LTS 1.424.3

[Kohsuke Kawaguchi]

Thanks for catching this. Respinning.

On 02/13/2012 01:56 PM, Vojtech Juranek wrote:

Hi,
during the testing I found out that auto-install of JDK doesn't work in most
of the cases (it works for jdk 1.7.0 and as I tested it in against 1.7.0 I
didn't spot it  in previous releases). I track down that it was logged as
JENKINS-11420 and fixed  quite long time ago.

It's marked as "major", but I would consider it as quite important feature and
therefore I backported the fix. Could you please re-spin? It passed all
automated tests [1].

Thanks
Vojta

[1] http://ci.jenkins-ci.org/view/Jenkins%20core/job/jenkins_lts_branch/43/


 Done. It's at their usual places: http://mirrors.jenkins-ci.org/





--
Kohsuke Kawaguchi  http://kohsuke.org/

Re: Backports for LTS 1.424.3

[Vojtech Juranek]

Hi,
during the testing I found out that auto-install of JDK doesn't work in most 
of the cases (it works for jdk 1.7.0 and as I tested it in against 1.7.0 I 
didn't spot it  in previous releases). I track down that it was logged as 
JENKINS-11420 and fixed  quite long time ago. 

It's marked as "major", but I would consider it as quite important feature and 
therefore I backported the fix. Could you please re-spin? It passed all 
automated tests [1].

Thanks
Vojta

[1] http://ci.jenkins-ci.org/view/Jenkins%20core/job/jenkins_lts_branch/43/

> Done. It's at their usual places: http://mirrors.jenkins-ci.org/

Re: Backports for LTS 1.424.3

[Kohsuke Kawaguchi]

Done. It's at their usual places: http://mirrors.jenkins-ci.org/

Re: Backports for LTS 1.424.3

[Vojtech Juranek]

Hi Kohsuke,
could you please post 1.424.3 RC for testing?
Final list of backports is as follows:

JENKINS-9519
JENKINS-11592
JENKINS-10831
JENKINS-11960
JENKINS-12340

and it passed all automated tests [1].
Thanks
Vojta

[1] http://ci.jenkins-ci.org/view/Jenkins%20core/job/jenkins_lts_branch/42/

Re: Backports for LTS 1.424.3

[Vojtech Juranek]

> https://issues.jenkins-ci.org/browse/JENKINS-10831

backported
Thanks!

> 
> On 28 January 2012 19:17, Vojtech Juranek  wrote:
> > Hi,
> > I found only this one:
> > JENKINS-9519
> > 
> > Any other proposals for backports into 1.424.3?
> > 
> > Thanks
> > Vojta