Remove the old README.md file was replaced by the README.rst. It's not
necessary to keep the file.
Change-Id: Idaa843e802c7c2d0f34429515d948d6fc05fb0a2
OpenStack docs are typically written in rst, so this should prevent
people from needing to learn another doc language to contribute to
the tripleo-ci docs.
Part of my motivation is that I'm planning to add some more
content here, and I'd rather not write a bunch of markdown that
would have to be translated to rst if we decide to do more formal
tripleo-ci docs (which we should).
Change-Id: I30f0477b4141b05a72ce1d5bb511488fc1ad867b
Once this matrix is moved to THT there is no
need to have it in tripleo-ci
Change-Id: Icd252236f2e25592c1c44520424b9f77b07fad3a
Depends-On: I9ef1acefc6e1f347528a48edcb4d997a9628fcf6
This scenario will only be ran for releases from Ocata and above;
Currently the extra service it deploys is Barbican, and is only
doing a key order from it, however, in the future we will test
encrypted volumes once we configure that in t-h-t.
Change-Id: I19ed2dddd88ae1752c6b7500c1790f801f6bfcd9
This reverts commit aa577a6e5c.
Since barbican is only in the overcloud in ocata, this has broken stable CI jobs for scenario002.
Change-Id: I452b5c9b0d87753cdb1b6a6cabe72e89000d11fe
Partial-Bug: #1640311
Add Swift to scenario002 and switch Glance backend to use Swift storage.
Note: scenario002 now looks very similar to regular multinode job, this
is going to change as we'll add more services in scenario002. Now we
keep the non-ha multinode job to be run all the time, since scenarios
are only run at demand when Zuul thinks we need it.
Change-Id: I29ed91bef3ee822453b064ee1fd28e4549a4e130
* scenario001 will test Telemetry with Aodh, Ceilometer, Gnocchi with file
backend, MongoDB and Redis.
* scenario002 will test Cinder with iscsi backend.
* scenario003 will test Sahara.
All jobs are experimental and non voting.
Change-Id: I8b1a3cfd44ed25eb121f9f2a0a8bc7316f681324
scenario001 is a proof-of-concept of how we can scale-up our CI by
spliting services on multiple CI jobs.
In the case of scenario001:
- Sahara is deployed.
- Cinder and Swift are not deployed.
- Glance uses file backend.
- Test to register an image in Sahara with the pingtest Heat template.
Also adds the testing matrix in README. The matrix will grow with more
documentation in the future.
Note for reviewers: we will have a pingtest per scenario. After some
testing I figured it was too much complex to try to maintain common bits
in a template and do nested stacks.
Change-Id: I25521a8bf8b91f7df4020de6599c4e8420e81fef
This updates the CI status report so that we generate
a tripleo-jobs.html-table that can be used to include
the CI status results in an external website. The
existing tripleo-jobs.html is still generated like
it was before.
Change-Id: I97285e147e79efbce3c16bf56e5a1eaa1656c1e4
Removes a bunch of the now deprecated toci_* files which now
exists in a different form under tripleo-incubator/scripts/devtest_*.
These scripts served us well over the past few months but the
time has come to focus this project solely on CI.
Well done old toci_*. Thanks for helping us raise the bar.
Change-Id: I4b9ec09e2c841c8b9374d21ea9f442cbe2ef53d3
Toci hasn't worked in nearly two weeks, ever since some stuff was refactored,
after this commit toci
1. is now back in sync with devtest in incubator.
2. brings up an overcloud by default (set TOCI_DO_OVERCLOUD=0 to stop a
undercloud).
3. clones all git repositories to /opt/toci (these arn't refreshed
between toci runs unless removed).
4. changes some env variables so that the default behaviour is now
closer to a what a user setting up a dev env would want.
5. removes unused patches and adds some new ones.
6. Adds a bit more documentation.
Change-Id: I19a34b1a400ec646f37310f265d1345bb6882dc3
Also, change the TOCI_IPMI_USER and TOCI_IPMI_PASSWORD options
to be TOCI_PM_ variables as well and also support multiple
users and passwords like the other options do.
Added a new option to control whether toci checks out the
various tripleO git repos it uses (incubator, diskimage_builder, etc.).
Also, internally reworks the git clone directory names we so they just
use the project names and not the GitHub org names as well.
These changes should facilitate easier integration within the upstream
jenkins.