The duplicated _version_ definition in the template prevented
the proper query of Chinese indexed content from Solr.
After the patch had been applied restart Solr and
reindex Chinese content:
sudo service jetty restart
cd /srv/askbot-sites/slot0/config
sudo python manage.py askbot_rebuild_index -l zh
Change-Id: Iaeabc73febc64c45996425225d3ee38db7f36f72
Right now we are using dib in hpcloud for devstack-trusty and
devstack-precise, this means we don't need the devstack-trusty-dib and
devstack-precise-dib labels and images to be provided by hpcloud. Remove
as much as we can while retaining the label without providers so that it
is there when we need to test dib images in rackspace.
Depends-On: I91f8d91318752e09a15d64c8b3c506cf83badca2
Change-Id: I7b82fb5162af908a86ffcb291d908a88867f3b06
Merge once things are back to normal with devstack-f20 in rax.
This reverts commit f22e31a3667006299187c7b5d51dfe0be1a3c2c7.
Depends-On: I0880c850dedada893d6e3e7e922c2994ece74930
Change-Id: Iff6cab82cb27d3f89c2af8223cad156d29a34d86
Generate a logging configuration template for nodepool that splits
image-build logs out into separate files.
A tool to generate and maintain this file based on the nodepool
configuration is added.
The logging configuration template this tool creates is added
alongside the main configuration. This is passed to the deployment
puppet module via a parameter.
Change-Id: I4723c68e5c83b536560e0a8bd7b73e312b46fc02
Depends-on: Iec77fb62c353d09186140b23912caf02225a643b
Nova can't boot our devstack-f20 images in rax-.( regions, so
temporarily drop them from that label and rely exclusively on
hpcloud to provide them for now.
Change-Id: I72bdcaf5d135726d7b5dadce30e4c944359b7eac
Currently the puppetmaster was hardcoded on a template.
Add the matching vars to the manifests that use them,
to make that setting configurable.
Change-Id: I2b641ec11284f325c22c242fabba26d0433bf85b
Add dib image-build for Fedora 21. This passes config validation in
I250ea4959740cfc4650e9057caba09ae7bc25768 and has fixed the issues
that lead to I2671e05e44b8a311f00dec024e67cbdc2864185c (there is no
need to specify the path to the image as done in that change, d-i-b
inbuilt paths + some redirects on the upstream server get you to the
right image now)
Change-Id: I197312214d6e849f2047363a51b1c71829982743
Apache 2.4 is the default version for Ubuntu Trusty.
Upgrade the apache configuration with the additional permissions
required to run on both Apache 2.2 and Apache 2.4.
Follows work done in I70d814a5eaefea5a3fb6a47bc215f4ab64be884f
Change-Id: Ie63cf936709b0f272b7129a2dc613d07a4d2fab2
Add ask.openstack.org to openstack-infra. Setup an all-in-one
askbot site based on existing deployment, including apache,
redis,apache solr,postgresql. See askbot.rst for further
details. Refactored to depend on vamsee's puppet solr module.
Depends-On: Iffe07d3a34087cb15151787bc683208425a27594
Change-Id: I36504eac7b953c3cce3e21a3559ac95b1bc12da7
Add released Fedora 21 image build to nodepool build. This is an
update of Ic6c21f715441a16a1d832163dc71921f25d654df
Note nodepool needs the fix below before it will work with this
Depends-on: I0ea4dc54edad7ef9b5d4ad59922f762ed8bbcd89
Change-Id: I5f013abbdd24f0f3edbb2cdc6d3a1f53a45827e5
The special py3k-precise nodes are no longer needed by any jobs, so
stop building them and clean up any custom Puppet related to them.
Change-Id: I40d5d09f28ef53583d239d2e852e9c50b5962cf8
Depends-On: Ie105674833e0a527f990a7448855a1090f5dc651
Rackspace no longer specifies CentOS point releases in their base image
names. This means no more CentOS 6.5 PVHVM image instead we have a
CentOS 6 PVHVM image. Go ahead and use the new non point release
specific image in nodepool to accomodate this change.
Change-Id: I3f9423722615fbacf5ee53a02987fc2052491ef5
Rackspace removed their "normal" non PVHVM base image option. Switch to
using PVHVM to accomodate this change.
Change-Id: I5fc4a40266306e14b733cffa1ed8567369fee7da
In 470f0401 I managed to set the hpcloud-b2 bare-precise build script to
prepare_node_devstack.sh which is the wrong build script for bare nodes.
Change it back to prepare_node_bare.sh so that everything is happy
again.
Change-Id: I96e2bce655dbf036358ee2c485ce549a33fc2b9f
Build hpcloud devstack precise and trusty nodes off of dib built images.
Last time we attempted similar we ran into bugs with nodepool not
handling labels with both snapshot and dib images. These issues should
be corrected now so that rax and hpcloud can have two different images
sources as configured by this change.
Change-Id: Ia6d67de14fbd681e946c0178f77c0a8e74c7dabc
We can build centos7 dib images with our new trusty nodepool server. Set
the min ready for that node type to 1 so that we exercise these images
by booting nodes off of them. Later we will want to add jobs to test
that these images can run tests successfully.
Change-Id: I970b13d1c35e652e31705e9971fa7e0a10fa046e
We have had connections regularly hang when communicating with hpcloud
api endpoints. These connections stick around for hours backing up all
nodepool work against that cloud. Mitigate this by setting an api
request timeout of one minute on the hpcloud provider managers.
Change-Id: Id7ec3da3c501adb8fdcf54b79b57a842dbe899bd
New qemu-img on trusty defaults to compat level 1.1 when building qcow2
images. This is a problem when clouds have older qemu-img versions that
only support 1.0 and older. Override the compat level and set it to 0.10
as suggested in qemu-img's manpage to ensure compatibility with clouds
when building nodepool disk image builder images.
Change-Id: I3edd209e9562a99e5efaea5fa59872a8e3460ef0
bare-f20 nodes are not used currently; there was some effort to get an
amqp unit-test working on them but I think that has moved in a
different direction. New testing should focus on f21
Change-Id: I67b2117057d647052e6a2f705c420b1eb0ec1435
Bring up Fedora 21 nodes on rackspace and hpcloud. Although we are
close to being able to build our own images
(Ic6c21f715441a16a1d832163dc71921f25d654df), things aren't quite there
yet.
In the mean time, let's use upstream F21 images for initial testing to
ensure we don't regress while it is being worked on.
I envisage that the existing F20 devstack jobs will disappear quickly
for this once we've got this stable.
Change-Id: If764f676f43c86830b4277adaaf71098e9d4140b
Increate max-servers for rh1 to 100. HW has been added and Quotas have also
been increased for instances, RAM, ports and floating ips to allow of the
increased usage.
Change-Id: Ief1d1402118bc4fb38229fa8bb06bbff42c66f00
Having environment configs on not-master breaks puppet apply. Also,
environments with puppet apply don't make sense.
Change-Id: Ief6c8e9bf7d71ddc63dba484390124546f14b93a
