Hi, Thanks for the insights. In the meantime, I've asked on Twitter for people to test "mvn clean verify": * Fedora 18 : rpm-1 failing https://gist.github.com/looztra/5891905 * centos 6.4 : rpm-1 failing https://gist.github.com/hgomez/5891894 * rhel 5.9 on ec2 : even worse ?!? https://gist.github.com/Batmat/5892567 * rhel 6.x on ec2 : same as 5.9
So to try keeping us moving forward, I propose the following: let's just remove/disable that rpm-1 test and try a release. If there's some regression in the go, then I just think we'll have to invest a bit on ITs so that they're less random anyway. What do you all think? Cheers Le 30 juin 2013 00:30, "Rickard von Essen" <rickard.von.es...@gmail.com> a écrit : > Hi, > > On CentOS 6.3 > rpm-1 failes with AssertionError: /usr/myusr/app/bin/filter.txt ... > > Last time I checked IT's was not running on Bamboo. > > I did some debugging of this some months ago and if I remember correctly > it seems like rpm cleans a up some files that the test is trying to check. > > A general problem with the integration tests is that they don't really > check the rpm sometime but some temporary files. > > Reagards > Rickard > > > > > On Sat, Jun 29, 2013 at 10:10 PM, Anders Hammar <and...@hammar.net> wrote: > >> I tried "mvn clean verify" on my MBP and IT #1 fails, but that's a know >> issue with mac and the library we're using for checking the rpm content. >> >> [INFO] Building: rpm-1/pom.xml >> [INFO] ..FAILED (6.9 s) >> [INFO] The post-build script did not succeed. OS name script missing, >> expected: name-Mac OS X.sh >> >> Here's my setup: >> Apache Maven 3.1.0-alpha-1 (c726cdd3a9ad5c3a419e1171f8c1925e336ead18; >> 2013-06-01 15:03:28+0200) >> Java version: 1.6.0_45, vendor: Apple Inc. >> >> The funny thing is that I could have sworn I installed JDK 1.7 the other >> day... >> >> /Anders >> >> >> On Sat, Jun 29, 2013 at 6:51 PM, Baptiste MATHUS <bmat...@batmat.net>wrote: >> >>> Hi all, >>> >>> As a followup, I tried to "clean verify" the rpm-maven-plugin with many >>> different ways since last week. >>> I've never been able to have all the ITs pass... >>> >>> The funny thing is that this is on my current Kubuntu and rpm apt-get'd >>> that I had the smallest number of failing IT (only one "rpm-1")... >>> >>> I tried at work with RHEL5 CI nodes. I just tried with an EC2 instance I >>> installed with a fresh 64 bits RHEL6.x. >>> There I get 6 errors on 7 ITs... >>> >>> Is there anyone that could just try running mvn clean verify inside the >>> rpm-maven-plugin (better with some system with rpm command, like fedora, >>> mandriva, centos, rhel and so on), and tell me if they're able to have the >>> IT passing. >>> >>> I'd really like not to have to give up with trying to handle this >>> release, but I'm a bit stuck. >>> Wondering if this is those ITs that are fragile and somehow dependent on >>> something restrictive (rpm command version or something), or if this is me >>> doing something wrong somewhere. >>> >>> What I checked: >>> * Bamboo-CI is OK. I'd be curious which system it's running. >>> * I tried with m221 and m305, doesn't change anything. >>> * RHEL5&RHEL6 KO for me. >>> * did only try with 64 bits. Might be something specific to 32 bits? >>> >>> Anyone having 5 minutes at least to check if they're able to clean >>> verify that plugin on their box? >>> >>> Thanks a lot. >>> >>> >>> 2013/6/26 Baptiste MATHUS <bmat...@batmat.net> >>> >>>> Hi Rickard, >>>> >>>> Thanks a lot for your efforts. >>>> Unfortunately, I was unable to apply the MRPM-141 patch, even if I svn >>>> up -r18277 before. >>>> It gives many conflicts. >>>> >>>> Note that I was able to apply the one of MRPM-137 without any issue. >>>> >>>> I'll have another look this evening I think. >>>> >>>> Cheers >>>> >>>> >>>> 2013/6/26 Rickard von Essen <rickard.von.es...@gmail.com> >>>> >>>>> You can ignore *MRPM-137*, it's unrelated. I have verified that I get >>>>> the same error as you on Ubuntu Server. With out debugging it in detail >>>>> I'm >>>>> quite sure it's the *org.codehaus.mojo.unix:unix-rpm* library that >>>>> fails to parse the output of *rpm -qlvp <RPM>. * >>>>> * >>>>> * >>>>> I rewrote the Integration Tests to only use groovy + rpm and not >>>>> depend on unix-rpm. Applying that patch >>>>> http://jira.codehaus.org/browse/MRPM-141 all tests passes also on >>>>> Debian based distros. >>>>> >>>>> // Rickard >>>>> >>>>> >>>>> On Tue, Jun 25, 2013 at 10:20 PM, Rickard von Essen < >>>>> rickard.von.es...@gmail.com> wrote: >>>>> >>>>>> Hi, >>>>>> >>>>>> I hade planed to do some work on the rpm plug. But currently I'm >>>>>> pulling two jobs so it hade to wait a while. >>>>>> >>>>>> I think http://jira.codehaus.org/browse/MRPM-137 might fix the IT >>>>>> for you. I use vagrant to verify that everything works in CentOS while >>>>>> developing on OS X. I attach my Vagrantfile. >>>>>> >>>>>> Happy hacking! >>>>>> >>>>>> // Rickard >>>>>> >>>>>> >>>>>> On Tue, Jun 25, 2013 at 8:07 PM, Baptiste MATHUS >>>>>> <bmat...@batmat.net>wrote: >>>>>> >>>>>>> That's exactly what I was trying to do :). >>>>>>> The thing is, I thought afterwards, I haven't an rpm-based system. >>>>>>> I'm running kubuntu. So I installed rpm through apt-get. I don't if >>>>>>> this is >>>>>>> known to have issues. >>>>>>> >>>>>>> FYI, in fact, I fear things aren't going to run totally smoothly >>>>>>> since I'm already running into an IT failure (rpm-1): >>>>>>> " post-build script did not succeed. /usr/myusr/app/bin/filter.txt >>>>>>> does not exist " >>>>>>> >>>>>>> I'm gonna have to dig a bit here. If necessary, I guess I'll install >>>>>>> some VM running RH or do it from some machine at work, but I'm not yet >>>>>>> even >>>>>>> sure the IT wouldn't fail there too. >>>>>>> >>>>>>> Cheers >>>>>>> >>>>>>> >>>>>>> >>>>>>> 2013/6/25 Mirko Friedenhagen <mfriedenha...@gmail.com> >>>>>>> >>>>>>>> Hello Baptiste, >>>>>>>> >>>>>>>> I think it would be great you made the release-manager. I would >>>>>>>> suggest a SNAPSHOT deploy of the trunk with the fix and Daniels >>>>>>>> confirmation, that the fix worked for him before, though. >>>>>>>> >>>>>>>> Regards Mirko >>>>>>>> -- >>>>>>>> Sent from my mobile >>>>>>>> On Jun 25, 2013 6:11 PM, "Baptiste Mathus" <bmat...@gmail.com> >>>>>>>> wrote: >>>>>>>> >>>>>>>>> Hi, >>>>>>>>> >>>>>>>>> No timeline. But though I personally dont use this plugin, if >>>>>>>>> nobody steps up/objects I can do it those days. >>>>>>>>> If the bug is already fixed, that seems quite an easy task to >>>>>>>>> provide users satisfaction. >>>>>>>>> >>>>>>>>> WDYT? >>>>>>>>> >>>>>>>>> Cheers >>>>>>>>> Le 25 juin 2013 17:54, "Daniel Lipofsky" <dlipof...@gmail.com> a >>>>>>>>> écrit : >>>>>>>>> >>>>>>>>>> 2.1-alpha-2 was released Aug 2012, and I know there is at least 1 >>>>>>>>>> bug fix since then that I am interested in, so I was wondering if >>>>>>>>>> there is >>>>>>>>>> a timeline for this. >>>>>>>>>> >>>>>>>>>> Thanks, >>>>>>>>>> Dan >>>>>>>>>> >>>>>>>>> >>>>>>> >>>>>>> >>>>>>> -- >>>>>>> Baptiste <Batmat> MATHUS - http://batmat.net >>>>>>> Sauvez un arbre, >>>>>>> Mangez un castor ! >>>>>>> >>>>>> >>>>>> >>>>> >>>> >>>> >>>> -- >>>> Baptiste <Batmat> MATHUS - http://batmat.net >>>> Sauvez un arbre, >>>> Mangez un castor ! >>>> >>> >>> >>> >>> -- >>> Baptiste <Batmat> MATHUS - http://batmat.net >>> Sauvez un arbre, >>> Mangez un castor ! >>> >> >> >
