summaryrefslogtreecommitdiffstats
path: root/roles/openqa
Commit message (Collapse)AuthorAgeFilesLines
* openqa/worker: add boot script to fix KVM perms and disable SMTAdam Williamson2017-09-153-0/+36
| | | | | | | | | | For some reason /dev/kvm has 0600 perms after boot on the ppc64 worker host. Also, qemu won't run unless SMT is turned off, on ppc64. I've just been doing this manually every time the box got restarted, but that's dumb, so let's make it happen on boot with a script and a service to run it. Signed-off-by: Adam Williamson <awilliam@redhat.com>
* Sigh. More JSON fixup.Adam Williamson2017-09-071-2/+2
| | | | Signed-off-by: Adam Williamson <awilliam@redhat.com>
* openqa/dispatcher: fix up images.json.ppc64 to be valid JSONAdam Williamson2017-09-071-38/+38
| | | | | | Sigh, JSON, why must you be so generally awful. Signed-off-by: Adam Williamson <awilliam@redhat.com>
* Whoops, == not =Adam Williamson2017-09-071-1/+1
| | | | Signed-off-by: Adam Williamson <awilliam@redhat.com>
* openqa/dispatcher: don't check out tests when locally modifiedAdam Williamson2017-09-071-1/+1
| | | | | | | Also don't check out the tests if we're on master branch, but have local modifications. Signed-off-by: Adam Williamson <awilliam@redhat.com>
* openqa/dispatcher: Install a variant images.json for stgAdam Williamson2017-09-072-0/+174
| | | | | | To get ppc64 tasks run on stg but not prod. Signed-off-by: Adam Williamson <awilliam@redhat.com>
* openqa/worker: correct WORKER_CLASS for ppc64 tap workersAdam Williamson2017-08-211-0/+4
|
* openqa/worker: more packages for createhdds, set SMT settingAdam Williamson2017-08-181-0/+20
| | | | | | One of the IBM guys helpfully pointed out we need to set this SMT thing to 'off' for VMs to run properly. I've no idea why, but this is how we do it!
* openqa/worker: couple of createhdds fixupsAdam Williamson2017-08-171-0/+4
|
* Add the createhdds cronjob file to worker roleAdam Williamson2017-08-171-0/+4
|
* openqa/worker: do createhdds on ppc workerAdam Williamson2017-08-172-0/+44
| | | | | This is needed to create the ppc virt-install images (has to be done on a machine of compatible arch).
* openqa/server: no root squash for rw access to factory shareAdam Williamson2017-08-171-1/+1
| | | | | Might be better to do this with some other account, but uh, for now...
* openqa/server: allow rw access to nfs share for ppc workersAdam Williamson2017-08-172-1/+2
| | | | | We need to create the ppc hard disk images from the ppc workers, so they need rw access to the share.
* openqa/worker: ed2k-ovmf doesn't exist on ppcAdam Williamson2017-07-251-1/+6
|
* Install git on openQA workers (RHBZ #1467086)Adam Williamson2017-07-251-0/+1
| | | | This will avoid some unnecessary error messages in logs.
* openqa/dispatcher: don't do git checkout when not on masterAdam Williamson2017-04-261-0/+16
| | | | | | | | | Like with openqa/server, sometimes we have the fedora_openqa checkout in some kind of testing state on staging. We want to skip the git checkout step when this is the case. This is just copied across from the server role. Signed-off-by: Adam Williamson <awilliam@redhat.com>
* openqa/server: *REALLY* fix the Apache status thingAdam Williamson2017-03-141-0/+6
| | | | OK, this version actually seems to work.
* Revert "openqa/server: try fixing the Apache status thing again"Adam Williamson2017-03-142-10/+3
| | | | | This reverts commit d6e32c9901c9f335a2fd2581dbaee1add9d1d9d0. It didn't work.
* openqa/server: try fixing the Apache status thing againAdam Williamson2017-03-142-3/+10
| | | | Seems like the config needs to be in the VirtualHost definition.
* openqa/server: Mostly revert previous commitAdam Williamson2017-03-141-11/+3
| | | | | | Actually openQA playbook already includes the apache role, which sets up the config file. The 404s are for some other reason. But keep the comment change.
* Enable mod_status on openQA server boxesAdam Williamson2017-03-141-3/+11
| | | | This should stop collectd generating tons of 404 requests.
* require resultsdb_conventions-fedoraAdam Williamson2017-03-081-1/+1
| | | | These roles need the -fedora subpackage, not just the main lib.
* openqa/server: check stdout not stderr for database creationAdam Williamson2017-03-071-1/+1
|
* openqa/server: allow createhdds to failAdam Williamson2017-03-071-0/+1
| | | | | | | | It does, sometimes, and that probably shouldn't stop the play. We likely already have some older images, and even if this is the *first* creation, we can go ahead with the rest of the deployment safely enough, and debug the image creation problem later.
* openqa/server: remove a no-longer-needed workaroundAdam Williamson2017-03-071-5/+1
|
* Use packaged resultsdb_conventions, not git installedAdam Williamson2017-03-071-0/+1
| | | | I'll wipe the from-source installs manually, seems easiest.
* Drop 'remove old manual install' steps that served their purposeAdam Williamson2017-03-061-11/+0
|
* Switch from git checkout to packaged install of openqa_clientAdam Williamson2017-03-061-19/+6
| | | | In check-compose and openqa/dispatcher roles, now it's packaged.
* openqa/server: enable update test restart plugin on prodAdam Williamson2017-03-041-3/+1
| | | | Seems to be working fine in staging, so let's do it.
* openqa/server: fix one of the git checkout conditionsAdam Williamson2017-03-041-1/+1
|
* openqa: install the 'update restarter' plugin, enable on stgAdam Williamson2017-03-012-2/+7
| | | | For now, just enable this on staging to check it works okay.
* openqa/server: load the update test templatesAdam Williamson2017-02-281-1/+6
| | | | Forgot these were in a separate file...
* openqa/dispatcher: update respin cronjob for new scheduler nameAdam Williamson2017-02-221-1/+1
| | | | | | The scheduler binary name changed from 'fedora-openqa-schedule' to 'fedora-openqa', but I forgot to update the call in this cronjob. Oops.
* openQA roles: use python2-fedfind not fedfindAdam Williamson2017-02-181-1/+1
| | | | I split the package, we don't need the CLI, only the lib.
* openqa/dispatcher: more config fixingAdam Williamson2017-02-151-0/+2
| | | | goddamnit, adam, why'd you make so many options.
* openqa/dispatcher: fix schedule.conf againAdam Williamson2017-02-141-1/+3
| | | | gah, section names.
* openqa/dispatcher: use 'localhost' as oqa hostname in infraAdam Williamson2017-02-141-0/+2
|
* openqa/dispatcher: tweak template logic a bitAdam Williamson2017-02-141-2/+4
|
* openqa: update roles for openqa_fedora_tools repo splitAdam Williamson2017-02-145-30/+43
| | | | | We split fedora_openqa and createhdds out into their own repos, this should adapt to the change (I hope).
* openqa/dispatcher: correct config for resultsdb reportsAdam Williamson2017-02-101-1/+6
| | | | | | | This is a bit sloppy and infra-specific ATM, but does the job. I have a PR in for openqa_fedora_tools which changes the report behaviour a bit, once that lands I'll clean this up somewhat using variables.
* openqa/dispatcher: correct openQA URL for resultsdb linksAdam Williamson2017-02-062-4/+10
|
* openqa/dispatcher: enable ResultsDB reporterAdam Williamson2017-02-021-0/+1
|
* openQA: install from stable for nowAdam Williamson2017-01-311-3/+6
| | | | | flippy flippy...u-t packages are untested, don't want them getting to stable.
* openQA: update name of default git branch for testsAdam Williamson2017-01-121-1/+1
| | | | Now it's master, not develop.
* Update some QA git repo locationsAdam Williamson2017-01-121-1/+1
| | | | We're moving stuff. Look out!
* openqa: handle 4.4-32 and 4.4-33 behaving differentlyAdam Williamson2016-12-201-2/+6
|
* openqa/scheduler: add a cron job to schedule live-respinsAdam Williamson2016-12-142-0/+11
| | | | | | | There are no fedmsg notifications for the live respin composes, so we just try scheduling them every hour; when we've already tested the current compose this will not create any new jobs, and when a new compose shows up, this will test it.
* Add a relvalconsumer role, run it on openQA servers for nowAdam Williamson2016-11-301-1/+1
| | | | | | | | | | | | | | | | relvalconsumer is the fedmsg consumer bot that creates Wikitcms release validation test events. Up till now it's just been running on one of my personal boxes; we should really move it to infra. Run it on the openQA servers for now, as there's nowhere more obviously correct, and I have root access there to fix problems. We add new inventory groups because it's *really important* that there be exactly one (no more, no less) production consumer at any given time. I don't want to just use the 'openqa' group for this because it's vaguely possible it could contain more than one host in future, and we really wouldn't want that to result in there being two production relvalconsumers running.
* In ansible 2.2 always_run is depreciated. Switch to check_mode.Kevin Fenzi2016-11-013-6/+6
|
* openqa/worker: use distro edk2 packageAdam Williamson2016-10-261-1/+1
|
generated by cgit (git 2.34.1) at 2025-09-26 21:24:24 +0000