These two jobs share the same volume, so use the common name as the
argument, otherwise we fail to release.
Also, avoid them running at the same time so one doesn't release the
volume while the other is still updating it.
Change-Id: Ied4feaec7fe5f0bfd459fa080b55b452e21416d4
The web related settings are relevant to zuul::web, not to zuul. Add
them here.
Change-Id: I7160826b9213f0c629374051f7c7043b553e00ed
Depends-On: https://review.openstack.org/557084
The referenced key has been added to hiera.
Change-Id: Id94db69498c462cb97c3721e96a645e5192a0b97
Depends-On: https://review.openstack.org/555968
Story: 2001382
Task: 6092
Because of ipv6 support in limestone-region one, we don't actually
need to ask for FIPs. Stop doing this so we can bring more servers
online.
Change-Id: I48fe176c4a78683834246f59850276133c26e19f
Signed-off-by: Paul Belanger <pabelanger@redhat.com>
We'd like to take advantage of the IRC logging that is managed by
openstack-infra. #softwarefactory is a project that packages
openstack-infra tooling:
https://softwarefactory-project.io/docs/
Change-Id: I3bd95eb06414ff7a27d6332591f65468e74bc5cd
Signed-off-by: Paul Belanger <pabelanger@redhat.com>
Copy/pasta from puppetmaster.o.o meant that we tried to put and use
the certificate in /etc/openstack/ which is not a thing on nodepool
nodes.
Fix that to be the right location and update the nodepool clouds.yaml.
Change-Id: Ib9c17a58fc2aa9ad5eb9091a6a3fd23ff5825731
This reverts commit 37c26280bfba775acfd1455be895ef6488efbbb5.
This was originally removed so that EPEL would not be installed in
our nodepool images. We no longer run this script when creating
those images, so it can be safely restored. It should be restored
because without this, we are unable to spin up a cgit server
(which requires epel).
Change-Id: I93e6487ab9a2eae5f927abf43aea7e3e19c5ac9e
Shade no longer uses novaclient. shade also strips links dicts from the
resources it returns. shade also now depends on openstacksdk, which does
not strip links dicts.
Change-Id: Ifb6a8280e548cb55932cae4a2bba8e1fa5b34c3c
When running the playbook, it's not immediately clear which task is
running without names. Add names. Also, update the whitespace to be more
in-line with how we write playbooks for zuul.
Change-Id: Ia189b8da6ded882aeb1fcff4932a1f9586027f80
Rather then creating per fqdn hiera entries for secrets, move to use a
group. This avoids the need to duplicate data.
Change-Id: I2208343b5281f70fc0850c0fe4e85038a53ed189
Signed-off-by: Paul Belanger <pabelanger@redhat.com>
This also removes review-dev.o.o, as it is an CNAME for
review-dev01.o.o.
Change-Id: I82624c5c8c7a5824fc925cd005dc1ed3bdf7c1bb
Signed-off-by: Paul Belanger <pabelanger@redhat.com>
This makes the credentials available so nodepool launcher and builder
can use them.
The limestone credentials have been added to hiera for the nodepool
group.
Change-Id: Idb56db19110e6b30f6231869ff278b90caf99f4b
Limestone provides both ipv4 and ipv6 addresses. Add the necessary
configuration to clouds.yaml to make it work.
Change-Id: I3c717132a42f44aaed6568de6780b1972cbf63d0
This will allow us to bootstrap a server with gerrit users, then
attach the volumes with hold the git repos for gerrit, then we can
remove this flag and properly puppet the rest of the server.
We also create a 2nd node in site.pp as we need both server to be
online for about 2 weeks, this is to give users enough time to make
firewall changes if needed for the new IP address
Related-to: I9159c941ece4f6928204601b9933d7a953baa2dd
Change-Id: I88826298818a690d4c98b60a9fbf444fba48cef6
Signed-off-by: Paul Belanger <pabelanger@redhat.com>
Now that volumes have been attached to review-dev01.o.o, allow puppet
to finish the gerrit configuration.
Change-Id: I392b26fcb3a787dc1a3022b0c3a6d0f9758526f6
Signed-off-by: Paul Belanger <pabelanger@redhat.com>
This will allow us to bootstrap a server with gerrit users, then
attach the volumes with hold the git repos for gerrit, then we can
remove this flag and properly puppet the rest of the server.
Change-Id: I9159c941ece4f6928204601b9933d7a953baa2dd
Signed-off-by: Paul Belanger <pabelanger@redhat.com>
Jeffrey Altman has pointed out that our settings are not optimal for
our use cases. Turning up threads and callbacks is a start. We
should evaluate the other settings too.
Add notes on how to apply settings manually
Change-Id: I1405b21f97c1ac2d3bd99ffbba18e5fd0ff959b1
A request from mnaser to use raw images for the provider. There is
also a proposed change to os-client-config to add this, but will need
to be tagged as a new release.
See: https://review.openstack.org/554351
We can revert once os-client-config has been released to pypi.
Change-Id: Ifc3d5656b7b972d5978b4d806c41bf2cc3443842
Signed-off-by: Paul Belanger <pabelanger@redhat.com>
Bring mirror02.ca-ymq-1.vexxhost.openstack.org to test if having no
bandwidth limits on our mirrors results is faster jobs.
Change-Id: If5efb199b1966a02ba28fd0cf309cc1c0dc05c02
Signed-off-by: Paul Belanger <pabelanger@redhat.com>
This site provides significantly higher bandwidth for
US based sites (compared on a Amazon EC2 US-EAST instance
it is 5MB/s vs 70MB/s)
Change-Id: Ia9e8dd40cfa08306812b8627c5843a0cb3c6f3b8
I had to figure out this command to restore some of our reprepro
mirrors, add notes on it.
Cleanup some other things I noticed; quotes around REPREPRO command,
clarify the situation with pointing mirrors at the RO/RW volumes and
refactor discussion.
Change-Id: I3e3f763929bf74e0c4faf29b13973dcf1df36975
