Changes include changing upstream git clone and pinning to the version
we run in production. Also new apache 2.4 syntax.
Change-Id: I3a214a42d3c49a1ea36fe84d21a275d736dfb889
Signed-off-by: Paul Belanger <pabelanger@redhat.com>
The ceph debian mirror does not have udeb components so don't attempt to
mirror them. Attempting to mirror them makes reprepro fail.
Change-Id: Ica8e99092d762445af78bb0a7f7f789c8576a1c5
mod_rewrite has inconsistent behavior inside of a directory tag. For
example, it doesn't appear to update the value of REQUEST_URI in
subsequent redirects, and the modifiable URI passed to the RewriteRule
is different from that passed to the RewriteCond.
This patch moves the rewrite rules for the npm mirror out of the
directory block, in order to fix unexpected 500 errors when accessing
packages with single file names. It also switches the <directory>
configuration block to only match on subdirectories of /npm/, so that
listing the contents of /npm does not redirect the user to the
/npm/index.json/ package.
Why that package exists is left as an exercise to the user.
Change-Id: Idbad41867d3d3546a59fe0c4fa54c5be22e31438
Same as the previous nodepool patch - it's important that we configure
internap correctly in our other clouds.yaml files.
Change-Id: I100fa7f7aa4836579dc8ddcfd40d78402aef38ef
It's not possible to determine which networks are public and private at
internap. But we can configure them.
Change-Id: Ica0d271188261c96e03a6e8af9b664b1031ab945
Mirror the trusty packages for the ceph hammer release to aid in process
of making ceph testing more robust. Use reprepro which is already in use
to mirror the main ubuntu trusty repos.
Change-Id: Ifd09272c7b1e07de9135be5a96be06153a3f611e
This patch makes the necessary modifications to host npm packages
on our mirrors. Changes include:
- index.json as a DirectoryIndex
- Rewrite rules for the AFS folder structure
- mod_substitute invocation to correctly reference absolute tarball
url's described in the index.json files.
Change-Id: I1a5d560068e37e20ec565b8d864f89732ff5ad74
This patch makes the necessary modifications to host npm packages
on our mirrors. Changes include:
- index.json as a DirectoryIndex
- Rewrite rules for the AFS folder structure
- mod_substitute invocation to correctly reference absolute tarball
url's described in the index.json files.
Change-Id: I7f7c6caa6af16101cf104f950968f3c6319b8243
This patch adds a cron job to our mirror_update servers which will
begin mirroring the npm registry into our own AFS mirror directories,
under the /npm path.
It uses two major components: First, the registry-static project,
available here: https://github.com/davglass/registry-static
Secondly, a custom implementation of the fs-blob-store named
afs-blob-store, which creates a virtual filesystem that automatically
translates paths to accomodate AFS' folder size limit. This project
is currently still hosted on github, however will be added to
OpenStack soon: https://github.com/krotscheck/js-afs-blob-store
This will require creating ACL's as well as the AFS principal before
this can be landed.
Change-Id: Ib78852fc5cde4fb04fe6439e01a591ca2781ce21
The TripleO cloud does not work when the neutron-based floating ip
logic creates a floating ip directly on a server port. The problem is
that the floating IP is correctly created and correctly associated with
the server, but Nova has no clue that this has happened, so GET calls to
/server do not show that the server has the floating IP. As that's the
way we track success, we can't use it. However, if a floating IP is
created using the nova proxy, it all works - so just configure the cloud
with "has_network: False" so that direct neutron interactions are not
attempted.
Change-Id: Idb90cad36feb99bc5aa19da4874ce97f868681c7
In an effor to better debug reprepo failures, lets increase the
logging.
Change-Id: I66073a26a60721221587b5e94825fdfdfe52fef9
Signed-off-by: Paul Belanger <pabelanger@redhat.com>
Otherwise, we get the following errors:
Error parsing /etc/reprepro/ubuntu/distributions, line 42: Second
appearance of 'Origin' in the same chunk!
Change-Id: I11fb815eddae1efa7ff072f8065af5b9ddee5e85
Signed-off-by: Paul Belanger <pabelanger@redhat.com>
Here we are removing devstack-trusty and adding ubuntu-xenial logs.
Change-Id: Ia4e2b3426442849efa02fe932d86d8ce5a1df48f
Signed-off-by: Paul Belanger <pabelanger@redhat.com>
The default of shade is not to cache server or port lists. However, we
would like to do so in nodepool - so configure that behavior.
Change-Id: Ic6a64b67a429910396dd801a4ef7577496e40fbf
When the extra OpenStack Health link was removed in
I293e326e45e1939dc57556d1477434cd1cfb56a2, the indices for the
'tabsName' array were not updated to reflect the change causing all
links after 'Rechecks' to be off by one. This patch restores the
correct index order to fix the issue.
Change-Id: I4eea976c3fb1a0b9b75d38ba5895b7d5fa5b0de3
We can use textwrap to break up the ridiculously long lines in the
logging configuration (with less images, it's less ridiculous than it
used to be, but it's still easier to parse). Only trick is to put
spaces between the arguments to give it somewhere to split (and better
for readability) and stop it splitting on words/hyphens.
Change-Id: I04840e98b0564dc531b74b96e8595f9d18b284af
This adds the centos7, debian-jessie and removes devstack-centos7-dib.
Change-Id: Ifcc8390b32713601fae98c96b761e313b7d2d6f3
Signed-off-by: Paul Belanger <pabelanger@redhat.com>
In Ubuntu, the archive configures backports as something you must
explicitly request when installing packages. One does this by specifying
a package name of 'backports/$packagename'. But our reprepro mirrors of
backports do not have this flag set, and so, all systems get backported
versions of packages, instead of the one from the released stable
archive.
Change-Id: I8f266dc7045e971567ed934d1206c7b83cacd178
The old logstash config was not ruby 1.8 compatible and we got funny
results out of it. This version should work with ruby 1.8 and beyond.
Change-Id: Ibe824dda7c96e5b333329ce25f65a14d3ebdef9c
Logstash 2.0 defaults to HTTP elasticsearch output which means that the
elasticsearch output (no _http) does HTTP and a new elasticsearch_java
output exists if you want to continue doing the native api output. We
had been doing HTTP so just need to update the output name. The host
parameter is also deprecated and you must pass an array to the hosts
parameter instead so update that as well.
Note that this switches from using a local ES daemon to talking to the
cluster itself directly because new logstash is able to load balance
over http. This reduces the overhead necessary to have resilient ES
connections.
Note this is not compatbile with Logstash 1.3.3 which is what we are
currently running so this change should only go in as part of an upgrade
to Logstash 2.0 and beyond.
Change-Id: I788ecb936f9fa5a006332ed626f90c33a255d9bf
Due to the possibility for multiple OSIC clouds we need to distinguish
between them in our clouds.yaml. Do that now before it becomes a problem
later and refer to the current cloud as osic-cloud1.
Change-Id: I3f35db2911a44200f0486e71fc215d021aa7c227
Add vexxhost account credentials to our various clouds.yaml files. This
covers the all clouds, ansible, and nodepool clouds.yaml files. With
this in place we can work to deploying tests onto vexxhost.
Change-Id: I42101e9acc9f62897a3f63b85dd34a14adcf2394
Project names are easier for humans to deal with use the project_name
key in clouds.yaml for OSIC not the project_id key.
Change-Id: I15b6424e355c711941a43e78116ffb71f6647cb7
This adds clouds.yaml information to our three clouds.yaml files for our
two users in the OSIC cloud. This will let us manage the OSIC cloud
resources and start deploying tests to OSIC with nodepool.
Change-Id: I5a392d165fb6db2e70036008a55cd99eed237ab4
In I9f61c4bb1f5634df0b0da909466b4c1b1b149ba5 when the Release link
was removed, a QA health link was added.
However, I57825f7e04bbdb405f7a1e67e4639d6afd7f3c5c is actually the
preferred change to this, which has the wording "OpenStack-Health"
instead. Matthew Treinish left a comment saying so when I tried
to revert it in Ia9355c3f4206ba089c19a51556bb5321e9df9b6b
Change-Id: I293e326e45e1939dc57556d1477434cd1cfb56a2
Ansible-clouds.yaml becomes /etc/openstack/clouds.yaml on the
puppetmaster and is used forr the ansible dynamic inventory. When a
cloud there does not respond, the ansible inventory fails completely.
Remove infracloudwest from all-clouds.yaml until it comes back.
Change-Id: I34d265a60f0a97f040b6703ab74c93a8fd0063af
The infra-cloud-west is currently unavailable due to being relocated
causing the inventory for ansible to not be generated:
Inventory script (/etc/ansible/hosts/openstack) had an execution error: Error fetching server list on openstackci-infracloud-west:RegionOne:
This means ansible and hence puppet fail to run.
For now remove infra-cloud-west from clouds.yaml. This should be added back
in one the cloud is turned back on.
Change-Id: I16fe54c3c040c273977b0cb2e51bf7f804832380