Since this image is going to be short lived, we don't want to add
ubuntu-precise AFS mirrors for ubuntu packages.
Change-Id: I0dc0e6bbb9f695584945056528cfb7e02793bdcd
Signed-off-by: Paul Belanger <pabelanger@redhat.com>
This reverts commit b4a933c18f4791d297a42a7d6f4fcff71b48b4a5.
Currently all our DIBs are failing to build, as a result we are
reverting stackviz until we can properly test.
Change-Id: Id336c1c8df7123c67f821a3a72da77c470a3a1f1
Signed-off-by: Paul Belanger <pabelanger@redhat.com>
In order to get Stackviz running against tempest gate run data, it
will need to be pulled in its pre-built form (i.e. with npm deps
installed) from the nodepool images. This change adds a couple steps
to build Stackviz so that it is in a ready state for retrieval and use
on the logs server.
A future devstack-gate patch will pull the stackviz directory from the
nodepool images onto the logs server. This nodepool patch is required
first in order to prepare stackviz for download. Installing the npm
modules takes a considerable amount of time (~5-10 mins), so it is
best that this work is done on the images before gate jobs are run.
Change-Id: I9bc50b8eba696264d2b97ed20a37f28f4768abb8
Reprepro now as local repos for xenial-updates and xenial-backports on
our mirrors. Lets use them now.
Change-Id: Ic1dbb9cf7e041ecce1a872386e0e3ce15757e261
Signed-off-by: Paul Belanger <pabelanger@redhat.com>
When cleaning up comments from lines of the form:
foo bar # this is a comment
You want to remove remove everything after the first comment character
not after the last comment character. This is because:
foo bar # this is # a comment
Should get everything after bar removed not everything after is.
Unfortunately when you use str.rfind() you remove everything after is
not everything after bar. Switch to regular trusty find to fix this.
Change-Id: I78aa6b51b5be03bd3b8ce7885415442171218977
Because reprepro will not mirror repositories with 0 packages,
xenial-backports, xenial-updates, xenial-security are missing from
AFS. Until we have them populated, we'll only configure APT to use
the xenial mirror.
Change-Id: Id4d788097ae4e1a5d0f13cb84a46fbb9ac9a839e
Signed-off-by: Paul Belanger <pabelanger@redhat.com>
When we removed the restrict memory grub stuff because hpcloud was no
longer in use we also removed the configuration updates that set the
grub timeout to 0. This can save many seconds per instance boot purely
waiting for grub to timeout and continue booting the host.
Change-Id: I2d79ec4ebd55f927d868e06c8f00cf50f34f3b92
At the end of gate jobs, we get warnings from zuul_swift_upload that
are not actionable for us.
Use requestsexceptions to silence these warnings. Add requestsexceptions
to the venv that is created for zuul_swift_upload to run in.
Example file
http://logs.openstack.org/48/298048/1/check/gate-ha-guide-tox-checkniceness/62e2d16/console.html
The example shows 6 warnings starting with:
/usr/zuul-swift-logs-env/local/lib/python2.7/site-packages/requests/packages/urllib3/util/ssl_.py:315:
SNIMissingWarning: An HTTPS request has been made, but the SNI (Subject
Name Indication) extension to TLS is not available on this platform.
This may cause the server to present an incorrect TLS certificate, which
can cause validation failures. For more information, see
https://urllib3.readthedocs.org/en/latest/security.html#snimissingwarning.
SNIMissingWarning
/usr/zuul-swift-logs-env/local/lib/python2.7/site-packages/requests/packages/urllib3/util/ssl_.py:120:
InsecurePlatformWarning: A true SSLContext object is not available. This
prevents urllib3 from configuring SSL appropriately and may cause
certain SSL connections to fail. For more information, see
https://urllib3.readthedocs.org/en/latest/security.html#insecureplatformwarning.
...
Change-Id: I02b4b6c7b426f3e9aa498941e4d75f67066d4d27
This reverts commit 468f17c4f6afac507dba9d4783c7da6fde24f656.
This isn't working on fedora-23 or centos7 slaves for some reason making
it very difficult to get those nodes ready. Revert in order to do more
testing across the platforms.
Change-Id: I634e784b71a2d92a745f8d9c89532e953f48b9cf
Update iptables rules to allow slaves to talk to themselves over their
known IP addresses. This should allow services to listen on public or
private addresses in addition to the loopback address and talk to those
services over these IPs.
Specifically this should address a problem with rabbitmq coming up on
the public IP but not being able to talk to it due to iptables.
Change-Id: If851eb8ce97ca0c0c113f0a924ec0916c993cf87
This patch adds Debian Jessie diskimage-builder image to nodepool.
debian-packageinstall element is based on debian-minimal and provides
additional required packages, change is introduced in
I9486b1356eac74272e79655c9024d86ee486628c
Depends-On: Ic39ba2b5ceb5018efb75742547b2babf80827e56
Depends-On: I7a661c8b5d4bb632eb32b5018c55474c1c66a4e8
Depends-On: I4a5a8352388de46e9e7713fa11951fcde27eaad9
Change-Id: I851cb47e844a72a430b578dd63d8b230ee6e3ed9
Co-Authored-By: Igor Belikov <ibelikov@mirantis.com>
We have attic and stackforge projects all of which are now inactive.
Because they are inactive we don't need to be caching these repos on
every test slave image.
Change-Id: I4df78c3d542758ce6159c195c1407f1d56f565a0
Now that we have per-region apt mirrors for ubuntu, we should use them.
As part of doing that we need to disable looking for i386 packages since
we do not mirror nor use those - and we need to tell APT to not check
the GPG signatures of our mirror archive since we're not signing it.
We don't need multiverse or restricted in the gate, and they also
contain non-free software. Remove them from our configuration.
Change-Id: I26f5fc5cec224cf4f4ea361e0f71cecbf320fbb0
This patch adds the wheel mirror location for this particular operating
system/region combo to pip.conf.
Change-Id: I700aaceda2b263c7b2e4ba167a52862d08598108
Depends-on: Ic9b069313c7a4e105cd086bdc4516c6c5826d53e
Pip performs a case-sensitive comparison on the trusted host. Just
make sure all our hostnames are lower case.
Change-Id: I63a66978a599629cb713c2de99618455f40ed0cc
The configure_mirror script has been updated to point our nodepool
slaves at mirror.<region>.<cloud>.openstack.org/pypi rather than
pypi.<region>.openstack.org.
Change-Id: I895b419e7755f8ec0e5fa594eebc8f9d79d333ba
Depends-on: I3d7629d80fe1e8762344852fab4a942bc476b706
Rather than trying to sed pip.conf, just write the whole thing.
This means we don't have to wait for image updates anymore. It's
not clear to me why we didn't just do this once we started seding
it.
Change-Id: Ie6743e2a8f95f95cea19a0dbe9fceaedbff8bf94
On OVH and Bluebox, the memory layout is such that there is still
significant (~900M) memory above the 8192M address. Increase the
limit to encompass that, which will bring these providers up to
approximately 8G, while increasing hpcloud (which is the actual
target of this restriction) to about 8.5G.
Change-Id: I5c121be55cadad13ad5807968f33b492f9b1e215
Since https://review.openstack.org/#/c/196252/ merged, nodepool
injects a sync;sleep 5 after running the scripts, so having it in the
scripts is a double sleep.
Change-Id: Ie7a9315e6de67625eca15b45f1662e4c63f8bc10
The release has happened, the migration worked, all's right with the
world.
This reverts commit 04eb36588d80dd77a9c692f53470b7d727567604.
Change-Id: I28a3e9cf3b420a6ccca20c52c4e67adc3f5710c5
This commit adds a temporary version cap on subunit2sql to be < 1.0.0.
The 1.0.0 release includes a very large database migration which will
be slow to execute. The python DB api from >=v1.0.0 will not work with
a database that doesn't have the updated schema. So while the migration
is running let's cap the version we install to prevent everything from
breaking while the migration is running. (which might take days)
Change-Id: Iab89beb5c7aba8b744a62f5063e513b72cab0ec2
The gate-neutron-dsvm-functional-py34 job needs Python.h, and it is
currently not available when running Python 3.
Change-Id: I00ce26864db682a4577dc5360519918007aa5499
Partial-Bug: #1500400
This commit creates a venv for installing os-testr which will enable
all test jobs to have access to the subunit2html utility which has
been moved to live inside the os-testr package instead of as a slave
script.
Change-Id: I2050b54eb2def10438764f3eeb55ecf9caa874dc
This help when reading log files from nodepool. Otherwise we see the
following in the log files:
[1;31mWarning: Config file /etc/puppet/hiera.yaml not found, using
Hiera defaults[0m
Change-Id: I3a865e5107e2749ed44c144539af49e311e0125f
Signed-off-by: Paul Belanger <pabelanger@redhat.com>
This reverts commit 71d10ea626f5a62d6b48133e19f469b19c9a7f14.
It was unnecessary once the offending wheel got removed from PyPI,
and this never worked anyway. It was preventing the ability to boot
new nodes, erroring thusly:
./configure_mirror.sh: line 25: /tmp/pip.conf: Permission denied
Deleting the images to which the change was applied returned
nodepool to being able to boot new nodes again.
Change-Id: I0841c6a5a26cf5be22e2d8fea861bdceb0393842
Remove jobs which test stable/icehouse branches of repos tagged
release:managed as the branch has reached end-of-life and is being
removed from those repos.
Change-Id: I88a44cfa84597012af7da0bd22de02dc2349b1fa
If we have booted with cloud-init, then this status directory is
populated. Remove it so when we boot snapshot images, they behave as
if on a fresh system.
Change-Id: Idc9ce01290b659e3239d30be847221447a8e5e84
hpcloud has started sending metadata to cloud-init to mount ephemeral
disks. This ends up writing a fstab entry for /dev/vdb.
---
$ curl http://169.254.169.254/2009-04-04/meta-data/block-device-mapping/ephemeral0
/dev/vdb
---
It's unclear why this just started happening, but it did.
devstack-gate later attempts to repartition this into swap & disk
space and mount it elsewhere. So remove this mount from fstab -- it
shouldn't come back because the next thing we do here is tell
cloud-init to not use the metadata source.
Change-Id: I3787d0f7e5139e891686ffbb2970e65d09f112b1
Cleanup every use of `` for subshells in the nodepool and tools
directory , replacing them with $(), and finally making the scripts
consistent.
Change-Id: I2b05cd20f9c9a30ab88f8db235aa81da93b1fad3
Some packages may depend on python-setuptools, which is not
installed and cannot be reinstalled on CentOS 6.x once yum has
erased them, so use --skip-broken to avoid aborting. Also on this
platform --downloadonly causes yum to return nonzero even when it
succeeds, so ignore its exit code.
Change-Id: Iaada39ae81e1e47fe9d0bedba80fd19e4e0e6f38
During scripted snapshot image builds we remove python-setuptools on
rhel/centos 6. This uninstalls cloud-init on hpcloud centos6 builds
because cloud-init depends on setuptools. We don't actually need
cloud-init in hpcloud because they use dhcp and we have hard coded ssh
keys so just don't bother setting cloud-init config if the config dir in
/etc does not exist.
One could argue that we would possibly want cloud-init to regen host
keys for us, but centos' sshd service should do that for us if the host
keys do not exist.
Change-Id: I96621b0ab1574eb8db0f4394877d3c1fc8208576
The --downloadonly option to yum is provided by the
yum-plugin-downloadonly package. This is merely a virtual package
satisfied by yum itself in newer releases, but an optional package
in older ones such as CentOS/RHEL 6.x. Install it just to be sure it
will work, since we use this to pre-cache RPMs on nodepool images.
Change-Id: I9e2e1605f3721c410180aa46a81b7b731d08503a
