Re: Git plugin: how does it decide to build instead of aborting?
> > What do you have in your "Branches to build" configuration? It would > appear that it's something that doesn't match any of the remote branches. > That was the right question to ask... and I got it to work. Branches to build: fork/${FORK_BRANCH} And I even threw in the Merge before build: Name of repository: upstream Branch to merge to: ${UPSTREAM_BRANCH} Now it works. Thank you. -- You received this message because you are subscribed to the Google Groups "Jenkins Users" group. To unsubscribe from this group and stop receiving emails from it, send an email to jenkinsci-users+unsubscr...@googlegroups.com. To view this discussion on the web visit https://groups.google.com/d/msgid/jenkinsci-users/fd2401b2-022c-4eca-bb51-533bba0d89c7%40googlegroups.com. For more options, visit https://groups.google.com/d/optout.
Git plugin: how does it decide to build instead of aborting?
In the big picture, I want to give my users a job where they specify both the source (fork) and destination (upstream) git repository, as well as the source and destination branch names, and I want jenkins to merge the source (repo, branch) into the destination (repo, branch), then build. The source repo is a fork of the upstream repo, necessarily. The git-plugin clones and fetches everything without any problem, but it errors out in the end. I am puzzled. Here is the configuration: Repo URL: ${UPSTREAM_REPO} Name: upstream Refspec: +refs/heads/${UPSTREAM_BRANCH}:refs/remotes/upstream/${UPSTREAM_BRANCH} Repo URL: ${FORK_REPO} Name: fork Refspec: +refs/heads/${FORK_BRANCH}:refs/remotes/fork/${FORK_BRANCH} Additional behaviours: Wipe out repository and force clone Check out to specific local branch: ${UPSTREAM_BRANCH} Merge before build Name of repository: upstream Branch to merge to: ${UPSTREAM_BRANCH} The the user supplies the four parameters: UPSTREAM_REPO=ssh://g...@stash.domain.com/path/project.git UPSTREAM_BRANCH=master FORK_REPO=ssh://g...@stash.domain.com/~user/project.git FORK_BRANCH=master The the user hits build, and the git-plugin ends with an error: Seen branch in repository fork/master Seen branch in repository upstream/master Seen 2 remote branches ERROR: Couldn't find any revision to build. Verify the repository and branch configuration for this job. Finished: Failure There are no errors before that in the log. The workspace contains the .git folder, with those branches not checked out - despite the fact that I asked for a specific local branch to be checked out: git branch -a remotes/fork/master remotes/upstream/master What causes the Git plugin to error out like that? What makes it decide that it should abort instead of proceeding with the build? I use Git 2.4.3, git-plugin 2.4.0, jenkins 1.609.1. Thanks, Martin -- You received this message because you are subscribed to the Google Groups "Jenkins Users" group. To unsubscribe from this group and stop receiving emails from it, send an email to jenkinsci-users+unsubscr...@googlegroups.com. To view this discussion on the web visit https://groups.google.com/d/msgid/jenkinsci-users/351b9b72-0eea-4cd6-80ee-ae8da2d0a1aa%40googlegroups.com. For more options, visit https://groups.google.com/d/optout.
build pipeline plugin links are all 404
The links on the build result page created by the build pipeline plugin are all broken and yield 404. The links look like this: http://host/1997eef2-85a8-407c-ac43-11bc67cb7403/job/jobName/ What is that hex number for? How do I fix these links? Thanks, Martin -- You received this message because you are subscribed to the Google Groups "Jenkins Users" group. To unsubscribe from this group and stop receiving emails from it, send an email to jenkinsci-users+unsubscr...@googlegroups.com. To view this discussion on the web visit https://groups.google.com/d/msgid/jenkinsci-users/5ebdd9f6-d3ed-4d28-b25c-e4741c0ff9ff%40googlegroups.com. For more options, visit https://groups.google.com/d/optout.
Re: jenkins support for parametrized test results
That's almost what I am looking for: each parameter is reported in a separate report. Is there a way to see the trend of a parametrized test over time, regardless of the parameter value? I mean the trend of the test, not the trend of each parameter value. Also, when I have say 50 parameter values for one test, is there a way to report on the test first, then and after clinking on the test, see the report for each parameter value? If you know of such way to report parametrized test results, please let me know. Thanks, Martin On Tuesday, May 5, 2015 at 3:46:48 PM UTC-4, Mark Waite wrote: > > > https://jenkins.ci.cloudbees.com/job/plugins/job/git-client-plugin/org.jenkins-ci.plugins$git-client/lastCompletedBuild/testReport/org.jenkinsci.plugins.gitclient/GitTest/ > > shows an example of the presentation of results from a parameterized test. > > Mark Waite > > On Tuesday, May 5, 2015 at 9:49:51 AM UTC-6, bl0ck3r wrote: >> >> Does jenkins support the JUnit XML format of JUnitParams for reporting >> tests? >> https://github.com/Pragmatists/JUnitParams/issues/38 >> >> If not, is there a Jenkins test report plugin able to report and track >> parametrized tests reports? The report format has to be as easy as >> JUnit XML as I write the reports myself (the existing test >> infrastructures do not work for my multi-language projects). >> >> Thanks, >> Martin >> > -- You received this message because you are subscribed to the Google Groups "Jenkins Users" group. To unsubscribe from this group and stop receiving emails from it, send an email to jenkinsci-users+unsubscr...@googlegroups.com. To view this discussion on the web visit https://groups.google.com/d/msgid/jenkinsci-users/5fb5c211-49da-48d3-80df-4c85cd67ea79%40googlegroups.com. For more options, visit https://groups.google.com/d/optout.
How to lookup Git Build Data from past builds?
Jenkins is the authoritative source of whether a Git commit SHA1 has passed or failed to build. How do I, several weeks after the build is done, interrogate Jenkins about a specific Git commit SHA1? Seems to me I have to scan a hole bunch of XML files until I find the right one. Does not feel like a fast lookup to me, more like a slow walk of all the XML files (O(N) rather than O(1)). Is the Audit to Database plugin a good idea to solve this? -- You received this message because you are subscribed to the Google Groups "Jenkins Users" group. To unsubscribe from this group and stop receiving emails from it, send an email to jenkinsci-users+unsubscr...@googlegroups.com. To view this discussion on the web visit https://groups.google.com/d/msgid/jenkinsci-users/fb2812d0-2c9d-4acb-ac2f-c917ccc1ebf7%40googlegroups.com. For more options, visit https://groups.google.com/d/optout.
Re: How to disable automatic scheduling of the Git Plugin
Yes, Mark was also suggesting that that as a probable cause, however, I think the problem is different. I decided to open a feature request, where I show that the Branch Specifier is set by a build parameter variable: https://issues.jenkins-ci.org/browse/JENKINS-26664 Martin > You might try leaving the "branches to build" field > > You need to put the specific branch in the box, that causes it to ignore > all other branches. > > > > It is because you have multiple branches matching the: > > Branches to build: Branch Specifier (blank for 'any'): > > -- You received this message because you are subscribed to the Google Groups "Jenkins Users" group. To unsubscribe from this group and stop receiving emails from it, send an email to jenkinsci-users+unsubscr...@googlegroups.com. To view this discussion on the web visit https://groups.google.com/d/msgid/jenkinsci-users/12f161b6-631d-45e7-83fa-a811b389b9f7%40googlegroups.com. For more options, visit https://groups.google.com/d/optout.
Re: How to disable automatic scheduling of the Git Plugin
I see that message ("Scheduling another build to catch up with project-integration") in the console. My use case is the same as yours: all my builds are on-demand. We definitely do not want any self-triggered builds. On Wednesday, January 28, 2015 at 1:40:43 PM UTC-5, Dunnigan, Terrence J wrote: > > We have plenty of parameterized builds that we invoke on-demand, by > clicking ‘Build with Parameters’. They use git and build a specific branch. > They do not poll and we don’t use any external triggers. > > > > I’ve never seen a message like the one you describe. Where do you get this > message? On the job configuration page? > > > > Terry > -- You received this message because you are subscribed to the Google Groups "Jenkins Users" group. To unsubscribe from this group and stop receiving emails from it, send an email to jenkinsci-users+unsubscr...@googlegroups.com. To view this discussion on the web visit https://groups.google.com/d/msgid/jenkinsci-users/ad4d91fc-7599-4265-9b8a-232b460d57a7%40googlegroups.com. For more options, visit https://groups.google.com/d/optout.
How to disable automatic scheduling of the Git Plugin
The Git Plugin tells me that it is "Scheduling another build to catch up with project-integration" I do not use polling, I do not use any triggers, I have a plain job triggered by hand. How do I turn off any and all of the automatic scheduling performed by the Git Plugin? Thanks, Martin -- You received this message because you are subscribed to the Google Groups "Jenkins Users" group. To unsubscribe from this group and stop receiving emails from it, send an email to jenkinsci-users+unsubscr...@googlegroups.com. To view this discussion on the web visit https://groups.google.com/d/msgid/jenkinsci-users/10864b38-38cf-40bb-8f63-b0f5e9196072%40googlegroups.com. For more options, visit https://groups.google.com/d/optout.
Re: git-plugin: can it do pull requests (no github)?
I do not understand how the Branch Specifier works when there are two repositories. I think I get it with one repo, but not with two. In the case of two repos, the plugin checks out the master even when I specify a branch. I wrote a short study on my experiment and on how I came to this conclusion: https://docs.google.com/document/d/1GKjD7GCvqUlHAo9Nmqzrlhi6LN8mv6ZnpuYSaTnU13g/edit?usp=sharing Can you help me understand? Thanks. -- You received this message because you are subscribed to the Google Groups "Jenkins Users" group. To unsubscribe from this group and stop receiving emails from it, send an email to jenkinsci-users+unsubscr...@googlegroups.com. To view this discussion on the web visit https://groups.google.com/d/msgid/jenkinsci-users/c4477a4f-9b1f-49b6-9176-2bd5949908b9%40googlegroups.com. For more options, visit https://groups.google.com/d/optout.
Re: git-plugin: can it do pull requests (no github)?
Thank Mark. I did not realize your were a maintainer on this plugin until I visited the wiki page. I have decided to learn the Git Plugin by experimenting with it, one baby step at a time. I am confused by the purpose of the refspec, and I have reopened https://issues.jenkins-ci.org/browse/JENKINS-14572. -- You received this message because you are subscribed to the Google Groups "Jenkins Users" group. To unsubscribe from this group and stop receiving emails from it, send an email to jenkinsci-users+unsubscr...@googlegroups.com. To view this discussion on the web visit https://groups.google.com/d/msgid/jenkinsci-users/c07e1903-30d0-4dec-9f6d-89f8fc209f6e%40googlegroups.com. For more options, visit https://groups.google.com/d/optout.
Re: git-plugin: can it do pull requests (no github)?
Thanks for the offer. Yes, I'd like to see a config.xml. That should help me understand how to use the git-plugin (I think I over estimated what it is meant to do). -- You received this message because you are subscribed to the Google Groups "Jenkins Users" group. To unsubscribe from this group and stop receiving emails from it, send an email to jenkinsci-users+unsubscr...@googlegroups.com. To view this discussion on the web visit https://groups.google.com/d/msgid/jenkinsci-users/83b12583-5a32-4e5d-8f24-b66f13322e83%40googlegroups.com. For more options, visit https://groups.google.com/d/optout.
git-plugin: can it do pull requests (no github)?
I was under the impression that by installing the git-plugin, I would be able to support pull requests between two custom local repositories (say file:// - no github here). By that I mean that the git-plugin would clone a "central" repo, fetch a branch from a "forked" repo, and merge them. Then it would let the core of the build run. Then in the post build, the git publisher would push the merge back to the "central" repo if the build was successful (with a tag maybe). I tried to make the git-plugin do that but it seems I need to write some parts by hand as I cannot get it to fetch from the forked repository. Am I mistaken about the git-plugin? Can it do all of this by itself? Please do not suggest github, it is not an option (behind a firewall). Thanks, Martin -- You received this message because you are subscribed to the Google Groups "Jenkins Users" group. To unsubscribe from this group and stop receiving emails from it, send an email to jenkinsci-users+unsubscr...@googlegroups.com. To view this discussion on the web visit https://groups.google.com/d/msgid/jenkinsci-users/3f47515d-6821-4836-8268-44ce0497eec1%40googlegroups.com. For more options, visit https://groups.google.com/d/optout.
Re: Scripting language to learn
Scripts are great for writing process specific glue-programs and point tools. I learned Bash from the internet, but after a while I bought the O'Reilly book "Learning the bash shell", it was worth very penny and this accelerated my learning: don't waste time, buy a book. I also often find answers I need in the Bash Guide on Greg's Wiki bit: http://mywiki.wooledge.org/BashGuide I have found Groovy quite fun to learn, esp since it can access the java libraries (argparse4j being one of my favourites). The biggest advantage of Groovy over Python is that it runs on the JVM, so it was just two downloads to get everything. Groovy automatically downloads the additional libraries it needs via imports and Grape, not need to get IT involved, no root access required. I still use Python 3 quite a bit, but only with the built-in libraries, since getting external libraries installed was hard for me on some legacy OSes. I still know enough Perl to fix old scripts, but I don't bother writing new scripts in it. Nowadays, for any problem that requires data structure manipulations (trees, lists), I start with Groovy (I have read part of the book "Programming Groovy 2" Venkat Subramaniam) . Last time I checked, Groovy performed really poorly with regular expressions, so when I have lots of them, I use Python instead (I have read O'Reilly Learning Python by Mark Lutz). -- You received this message because you are subscribed to the Google Groups "Jenkins Users" group. To unsubscribe from this group and stop receiving emails from it, send an email to jenkinsci-users+unsubscr...@googlegroups.com. To view this discussion on the web visit https://groups.google.com/d/msgid/jenkinsci-users/31e43c13-4f29-4e30-ac20-4d8f5c2e9651%40googlegroups.com. For more options, visit https://groups.google.com/d/optout.
Re: Update center is missing a plugin but it is listed in default.json
Is there a logger that I can turn on which would give me more information on this problem? The log reveals nothing relevant. -- You received this message because you are subscribed to the Google Groups "Jenkins Users" group. To unsubscribe from this group and stop receiving emails from it, send an email to jenkinsci-users+unsubscr...@googlegroups.com. To view this discussion on the web visit https://groups.google.com/d/msgid/jenkinsci-users/935b30d7-2101-4654-9174-48ad0af3ec43%40googlegroups.com. For more options, visit https://groups.google.com/d/optout.
Re: Plugin to display arbitrary results?
Arbitrary data in tabular form, try: https://wiki.jenkins-ci.org/display/JENKINS/Summary+Display+Plugin For plotting data over time, try: https://wiki.jenkins-ci.org/display/JENKINS/Plot+Plugin -- You received this message because you are subscribed to the Google Groups "Jenkins Users" group. To unsubscribe from this group and stop receiving emails from it, send an email to jenkinsci-users+unsubscr...@googlegroups.com. To view this discussion on the web visit https://groups.google.com/d/msgid/jenkinsci-users/5b94c0af-96f3-46ef-9089-02266b161a8f%40googlegroups.com. For more options, visit https://groups.google.com/d/optout.
System groovy script: how do I import a custom class from the slave workspace?
Hi, How does one import a custom class from within a system groovy script, when this class is in a file in the workspace? The workspace is obtained like this: def thr = Thread.currentThread(); def currentBuild = thr?.executable; def workspace = currentBuild.getModuleRoot().absolutize().toString(); Then the class is loaded like this: String roots = [workspace+"/path/to/file"] def engine = new GroovyScriptEngine(roots) def JDU = engine.loadScriptByName('JenkinsDescriptionUtils.groovy') Then jenkins crashes with: FATAL: Cannot open URL: file:/path/to/jenkins/%5B/path/to/slaves/myhost/workspace/job-name/path/to/file%5DJenkinsDescriptionUtils.groovy I looks like Jenkins has its own classloader and prepends its own path to the path I want, thus breaking it. I tried the simple solution, that is using a plain import statement, but that too did not work. Thanks, Martin -- You received this message because you are subscribed to the Google Groups "Jenkins Users" group. To unsubscribe from this group and stop receiving emails from it, send an email to jenkinsci-users+unsubscr...@googlegroups.com. For more options, visit https://groups.google.com/d/optout.
Re: Loading job configuration pages is excruciatingly slow (minutes)
Thanks for the advice. 1) scrolling to the end while config is loading: it gets to the end (the last post-build step which is editable email notification) very quickly. so it seems it's busy doing something else 2) examining the threadDump while loading the config is hung at the bottom: nothing stands out in terms of plugin name. However, reloading the threadDump repeatedly while the other tab is slowly loading the configuration page show that there is a lot of activity going on. I have pasted the threadDump here http://pastebin.com/F2Yhkjmp (but it changes all the time) Does anyone see anything abnormal? -- You received this message because you are subscribed to the Google Groups "Jenkins Users" group. To unsubscribe from this group and stop receiving emails from it, send an email to jenkinsci-users+unsubscr...@googlegroups.com. For more options, visit https://groups.google.com/groups/opt_out.
Re: Loading job configuration pages is excruciatingly slow (minutes)
I should add that the CPU is railed all the time and that I use Jenkins 1.509.2. There is nothing else on that host but Jenkins. We have about 50 jobs, and thousands of results (and hundreds/thousands of JUnit xml files per build). Have I exceeded Jenkins capacity? How do I discover what Jenkins is doing, we're only running two jobs on two slaves, the master should have nothing to do but wait. top - 09:24:57 up 72 days, 8 min, 2 users, load average: 1.06, 1.18, 1.14 Tasks: 104 total, 1 running, 103 sleeping, 0 stopped, 0 zombie Cpu0 : 99.0%us, 0.7%sy, 0.0%ni, 0.0%id, 0.0%wa, 0.0%hi, 0.3%si, 0.0%st Mem: 4059656k total, 4022352k used,37304k free, 268708k buffers Swap: 2097148k total,0k used, 2097148k free, 2728332k cached PID USER PR NI VIRT RES SHR S %CPU %MEMTIME+ COMMAND 1627 jenkins- 20 0 1483m 656m 8936 S 99.9 16.6 5125456h java -- You received this message because you are subscribed to the Google Groups "Jenkins Users" group. To unsubscribe from this group and stop receiving emails from it, send an email to jenkinsci-users+unsubscr...@googlegroups.com. For more options, visit https://groups.google.com/groups/opt_out.
Loading job configuration pages is excruciatingly slow (minutes)
Everything in Jenkins is reasonably fast, but loading job configuration pages is frustratingly slow (minutes). Using Firefox, hosted on linux. Started happening a few months ago. I don't get it since Jenkins should only be loading the config.xml for the job... what else does it do that takes so long? -- You received this message because you are subscribed to the Google Groups "Jenkins Users" group. To unsubscribe from this group and stop receiving emails from it, send an email to jenkinsci-users+unsubscr...@googlegroups.com. For more options, visit https://groups.google.com/groups/opt_out.
HTML Publisher plugin no longer listed
The HTML Publisher plugin is no longer listed on the following page... can some one bring it back please? https://wiki.jenkins-ci.org/display/JENKINS/Plugins Thanks. -- You received this message because you are subscribed to the Google Groups "Jenkins Users" group. To unsubscribe from this group and stop receiving emails from it, send an email to jenkinsci-users+unsubscr...@googlegroups.com. For more options, visit https://groups.google.com/groups/opt_out.
Duplicate test in JUnit XML breaks test result links
The second link in the "All Failed Tests" table breaks when this JUnit XML is used, is it a bug? XML: failed with seed value of 123 failed with seed value of 456 It looks like the problem is that although Jenkins is able to link to the test results themselves, it does not create two test result instances in the javascript: >>> >>> Pay attention to the SECOND argument to the javascript:showStackTrace() call: test1 and test1_2 are different, while the FIRST argument is the same in both cases. Question: Can duplicate tests be accommodated somehow? If not, is there a test report format that supports running multiple iterations of the same test? Thanks. -- You received this message because you are subscribed to the Google Groups "Jenkins Users" group. To unsubscribe from this group and stop receiving emails from it, send an email to jenkinsci-users+unsubscr...@googlegroups.com. For more options, visit https://groups.google.com/groups/opt_out.
Re: End-user debugging Jenkins/Git Plugin
I tried the suggested Loggers and I see nothing in the logs. How do I find the right Loggers? -- You received this message because you are subscribed to the Google Groups "Jenkins Users" group. To unsubscribe from this group and stop receiving emails from it, send an email to jenkinsci-users+unsubscr...@googlegroups.com. For more options, visit https://groups.google.com/groups/opt_out.
Re: Looking for XML report format that supports multiple iterations per test
I am looking at the NUnit samples, but I don't find an example of how to write a test report. However, I have noticed the following with JUnix XML in Jenkins: if there are two identical tests in the test suite, Jenkins reports them separately: I wonder if this would work, or if it would silently break something else? On Thursday, May 23, 2013 9:29:31 PM UTC-4, Mark Waite wrote: > > The NUnit framework supports that type of execution ( > http://nunit.org/?p=testCase&r=2.5.5), so you might try the NUnit plugin > to see if its format meets your need. > > Mark Waite > -- You received this message because you are subscribed to the Google Groups "Jenkins Users" group. To unsubscribe from this group and stop receiving emails from it, send an email to jenkinsci-users+unsubscr...@googlegroups.com. For more options, visit https://groups.google.com/groups/opt_out.
Looking for XML report format that supports multiple iterations per test
Hi, I am looking for any XML report format publishable by Jenkins, that is able to report multiple iterations of the same test when the same test is executed with different random seeds. I am NOT looking for a test framework, nor a test runner. I am only looking for a test report format digestible by Jenkins for writing custom test results myself. I already use JUnit XML, but it is limited to the "package.class.testname" hierarchy, and does not appear to support multiple iterations of the same test with different seeds. Any pointers? Thanks. -- You received this message because you are subscribed to the Google Groups "Jenkins Users" group. To unsubscribe from this group and stop receiving emails from it, send an email to jenkinsci-users+unsubscr...@googlegroups.com. For more options, visit https://groups.google.com/groups/opt_out.
Why is this unstable rather than failed?
This JUnit XML file causes the build to be unstable. Why isn't it failed instead? SEED=1 SEED=2 Thanks. -- You received this message because you are subscribed to the Google Groups "Jenkins Users" group. To unsubscribe from this group and stop receiving emails from it, send an email to jenkinsci-users+unsubscr...@googlegroups.com. For more options, visit https://groups.google.com/groups/opt_out.
Why is SIGTERM trap in freestyle shell script not executed?
In a freestyle job, the build phase has the following script: #!/bin/bash trap 'echo "Death by SIGTERM"; kill $pid; exit 15;' SIGTERM sleep 10 & pid=$! echo "Waiting" wait $pid echo "Exit status: $?" When this is aborted with the UI [x] red box, the message "Death by SIGTERM" never appears, but the exit code reflects the abort by printing "Exit status: 143" Why is the trap content not executed? Thanks. -- You received this message because you are subscribed to the Google Groups "Jenkins Users" group. To unsubscribe from this group and stop receiving emails from it, send an email to jenkinsci-users+unsubscr...@googlegroups.com. For more options, visit https://groups.google.com/groups/opt_out.
Re: Looking for good Linux bash script examples
Bash is very good when the problem is solved by running unix commands. Learning bash is key, and I strongly recommend the O'Reilly book "Learning the bash shell". The bash shell is mature and even if the book does not cover bash 4, all the fundamentals are there. I can recommend three sites for bash: http://mywiki.wooledge.org/EnglishFrontPage http://wiki.bash-hackers.org/doku.php http://tldp.org/LDP/abs/html/ Exit codes are easy to master. An exit code of zero means the command has "passed", anything else means the command has failed somehow. For example: make my_target if [[ $? -ne 0 ]]; then ret=1 echo "Make has failed" >&2 fi I keep all my problems in the pass/fail category, and I never try to interpret non-zero exit codes because it gets too complicated. However it is key that any script underneath returns an exit code to its caller, all the way up to Jenkins, so make sure your Jenkins build step has an exit statement with an exit code that represents the "worst" error reported underneath. If you use pipes, you should always use "set -o pipefail". This will make sure that the pipe will not hide a failure (see man bash): $ (exit 1 | echo "hi") hi $ echo $? 0 $ set -o pipefail $ (exit 1 | echo "hi") hi $ echo $? 1 Now regarding Jenkins, when you have a parametrized build, Jenkins exports those parameters as shell variables that match your parameter names. You can see them in the console by adding the "env" command to your free style build. Also I recommend you always use the shebang to start your free style build (Jenkins will run your script with the -e option, but there are too many exceptions to "set -e", so I just use #!/bin/bash, see: http://mywiki.wooledge.org/BashFAQ/105) If you background jobs, you will need to keep track of their process ids, for example set a trap: function onKill { kill $pid ... other clean up you might want to do goes here } ./my_script & pid=$! trap 'onKill; exit 1;' SIGINT SIGTERM wait $pid if [[ $? -ne 0 ]]; echo "my_script failed" &2 fi You will also need to generate JUnit XML reports in the format that Jenkins understand. Java is not the language of my builds, and it was very painful to figure out what Jenkins supported. It was a lot of trial an error, and there are many unclear explanations on the internet. I ended up with something similar to this: http://pzolee.blogs.balabit.com/2012/11/jenkins-vs-junit-xml-format/ My bash build step does not generate XML reports, but commands inside GNU Make, which is called by my script, do (it is actually many layers below the Jenkins "script window" in my case). Last advice, keep the code in the build phase window to the bare minimum: put all the smarts in a script that Jenkins calls instead. This way you don't have to copy and paste bug fixes when you have the same script in multiple builds: you simply fix the script that Jenkins calls. Bash is very powerful if you learn it well. If you need to manipulate data structures, then you will need to move up to a higher level language like python, but if all your build does is call commands and watch exit codes, bash will do the job really well. Good luck. -- You received this message because you are subscribed to the Google Groups "Jenkins Users" group. To unsubscribe from this group and stop receiving emails from it, send an email to jenkinsci-users+unsubscr...@googlegroups.com. For more options, visit https://groups.google.com/groups/opt_out.
envinject plugin evaluated groovy script: how do I print text to the console?
I tried: System.out.println("some text") println("some text") manager.listener.logger.println("some text") and none of them work. How do I print text to the console from the evaluated groovy script of the envinject plugin? Thanks. -- You received this message because you are subscribed to the Google Groups "Jenkins Users" group. To unsubscribe from this group and stop receiving emails from it, send an email to jenkinsci-users+unsubscr...@googlegroups.com. For more options, visit https://groups.google.com/groups/opt_out.
Promotion blocked by current build: Waiting to acquire the workspace
Hi, My manual promotion process is being blocked and the thread dump shows this message: Waiting to acquire /.../workspace/project-name : Executor #2 for hostname : executing project-name » promotion » Promotion #40 This is terrible since the build I need to promote is actually an older build, so I see no reason why it should wait for the current build to finish. I don't even use the workspace in the promotion process because of the risk of a race with the current build. My builds take several hours, and I cannot afford to have a promotion process that forces people to wait. Obviously, the build promotion plugin was built assuming all builds are very short. Making the build shorter is not possible: synthesizing and simulating large chips takes a long time no matter what. Other than the build promotion plugin, what do you people use to promote builds? Cheers, Martin
Re: Build Promotion and Copy Artifact: current build is overwritten
On Friday, August 17, 2012 5:56:29 PM UTC-4, qazwart wrote: > > > On Aug 17, 2012, at 5:12 PM, bl0ck3r > > wrote: > > > Well, I think what I am trying to achieve is simple: promote a build > while another build is running on the same project > > Wait a second. Is the problem your copying in files to a project while a > build is running or copying files out of a build while a second build on > that project is running? > What confused me is that the "Copy Artifact" (which is recommended by the Promoted Builds Plugin as the solution for not interfering with the workspace) does offer as a default value, the workspace as the target directory for the copy of the archived artifact. When I changed the target directory to be an absolute path outside of the workspace, it worked. This was counter intuitive to me. > > Anything you want to save should be "archived" which is a post build step. > Archived artifacts are kept for each build, so they're not overwritten when > a new build starts. You can copy out archived artifacts w/o affecting the > currently running build. > Only if I the target directory for the copy is NOT the workspace. Martin
Re: Build Promotion and Copy Artifact: current build is overwritten
> > Another possibility is to copy the artifacts not to the build > directory itself, but to a directory outside of Jenkins (which is > pretty much what the Maven repository does for us). > > Well, I think what I am trying to achieve is simple: promote a build while another build is running on the same project. The artifact names are the same at each run, so the file names collide on the file system in the workspace during the promotion. Note that I clean up the workspace at the beginning of each new run of the job with rm -rf * because I want a clean slate when the job begins. Like you said, to solve the problem, I cannot copy the promoted build artifact into the current workspace, I need to copy it to an absolute path elsewhere. This is what I have done. It is an annoying workaround, I wish Jenkins promoted builds had their own workspace. BTW I am not using Jenkins for a Java flow, so Maven does not apply to my case (and nor does any open source flow for that matter). I am happy that bare bones Jenkins is development flow agnostic and I hope it stays this way. Cheers, Martin
Re: Displaying user in the Build History column on the left
I already use the description setter to set the description to the actual description! How to concatenate the name of the person launching the build to this description? My regex is DESCRIPTION=(.*)
Re: Displaying user in the Build History column on the left
Looks like I have to write my own code for this to work... this rules out this option for me for now. On May 15, 6:53 pm, Ken Bertelson wrote: > I use the Groovy Postbuild Plugin to do this. You can display custom icons > in the Build History, too.
Re: Displaying user in the Build History column on the left
I use this plugin: https://wiki.jenkins-ci.org/display/JENKINS/Description+Setter+Plugin I looked under the Jenkins configuration and the job configuration for "started by" but I could not find it. On May 15, 10:46 am, Peter wrote: > Use the description setter plugin. You need to set the description > with the text behind "started by".
Is it called "Project" or is it called "Job"?
The UI needs to be more consistent. In the main Jenkins menu, I can create a "New Job" called Foo, but after it is created, the page is titled "Project Foo". This is inconsistent and got me confused for a couple of days as I was learning Jenkins.
Displaying user in the Build History column on the left
How do I display the username of the person who launched the build in the Build History (that is the box on the left-hand side of the job page)
Displaying test results before job is finished
I have hundreds of tests in a job, and it can take several hours to complete. Is there a way to have Jenkins display the test results as the job progresses? Do not tell me to split the job into smaller jobs, it is not practical as I would end up with hundreds of jobs and this would be unmanageable. I prefer to have one longer running job with hundreds of small tests, than hundreds of jobs with fewer tests. Thanks.
ChangeLog parser without SCM
Is it possible to have a ChangeLog parser without developing the full SCM plugin?
Re: how to report on multiple iterations of the same test
I was hoping to have an additional column for the seed value. I do not know what a "TestRunner" is, and unless it works with SystemVerilog, I can't use it. On Mar 29, 3:27 am, Ullrich Hafner wrote: > Can't you change the name of the test case dynamically? E.g., using a > TestRunner? > > Ulli >
how to report on multiple iterations of the same test
I need to report on multiple iterations of the same test when it is run with different seeds. With the following report, jenkins does not see multiple iterations... how do I get it to report all the iterations? Thanks.