Remove unused functions from common_functions.sh after
Ie230aa0640e41a33e10e6656b69e1db4a6819127 merged. This commit deletes
all the functions that aren't referenced anywhere. Most of them are now
implemented in quickstart.
Also remove some of the scripts that were only used in these functions.
There is probably more cleanup to do.
Change-Id: I4fef5f8482f5eb7d4276b1d2d2563fd57d7419fb
Add missing bits to support stable/pike branch in TripleO;
Some adjustments might be done later but this is a first iteration.
Change-Id: I408a6b116b611b67b67f98e1bf7bc533ce22435c
Several love fixes needed here:
- We forgot to override STABLE_RELEASE so
the upgrade was failing because it was
deploying repos from Pike instead of Ocata.
- Allow only to check that Ironic API is running
on Newton or Ocata.
In master we can check that ipmi driver is
available.
- On old relases like Newton or Ocata we will
use the old heat calls on get host info.
Change-Id: I558e86f5cec139cef165b0908bb3861434c7c8bc
Depends-On: Ie79dfff667437305ff4109da5a9c42ae358b1a3e
The output from these commands could also be causing postci.txt to be
22MB (uncompressed), so switching to these commands may help with
Change-Id: I0066da26788e1bcf4d112dc3e14c26d0ef76f298
Related-Bug: #1698172
This will be useful to understand which workflow has been executed
by service_workflow_tasks added by change
I36a642fbc2076ad9e4a10ffc56d6d16f3ed6f27
Change-Id: I2fbc9968e27fae9e1384d5393558c06ddb1b5009
A couple of small changes intended to make it easier to debug
failures in the gate. First, run nova service-list and
neutron agent-list after failures to create the ping test stack,
not just after a failure to ping the vm. A service outage could
also cause a stack create failure.
Second, sort the rpm -qa output for easier comparing between
different runs of a job. Otherwise it seems to come out in a
semi-random order that doesn't diff nicely.
Change-Id: I9e3319811054668289193b24fd11ce6f0775ade9
Mitaka and earlier don't have openstack stack failures list,
so we're not getting any useful output in postci when those jobs
fail. If that command fails, try the old manual method instead.
Change-Id: I23b7f6eb3a8c94c95448fe27319c5c950285fd1d
The new 'openstack stack failures list' will help us to track down what
went wrong in overcloud deployment, update or delete.
Change-Id: Ie536334990ad933776b53b73b5439315c15b0637
Being able to set a different $TRIPLEO_ROOT other than /opt/stack/new
makes the toci scripts more useful when running locally. There are also
lots of places where /opt/stack/new was used directly, when we
could/should be using $TRIPLEO_ROOT.
Change-Id: Iac212866fc343f3f3f930fad57b3a52236513d34
If a run fails before the overcloud is deployed or we are running
get_host_info.sh after deleting the overcloud then all of the
commands that operate on the stack will fail. Let's not even run
those if we don't find the stack.
Change-Id: I06ff879424793ea1372889ae2c0e0740c45cb5c8
Being able to see all of the facter facts can help with debugging puppet
issues. This adds ~3s to get_host_info.sh.
Change-Id: I8757602047696e40c7412046ca497492760ddad3
This way we can track how they change over time, in addition to
seeing how they behave in a single run.
Change-Id: I28618906e119ad22d10d087e2faf75337849cb74
Looking through our graphite metrics, by far the longest time is
spent deploying the overcloud. This change allows us to collect
more granular data about what is using that time in the deployment.
At some point we may want to add this to the graphite data so we
can track it over time, but just making it available in the postci
logs is a good first step.
Change-Id: Ibae0f6259146dc0da73d43aba0ff31a24ae71eff
We get some intermittent failures when trying to ping default gateway on
IPv6, it might be interesting to compare routes on successful
vs. unsuccessful deployments. And having `ip route` output could be
useful for some other debugging in the future too.
Change-Id: I0380e87282f6a65622eafa88f125e7f137f63362
Right now the ps command we run in get_host_info.sh doesn't include
things like how much memory a process is using. This change adds
some more switches to ps so it will include that information.
Change-Id: I884bb36d48cc5584d4d845351aa94fef391f2ffe
The script collecting host information doesn't run as root, we need sudo
to make pacemaker status queries work. Also adding crm_verify output.
Change-Id: I5858ddf9d1d317591b4ec4e6d7633115b636709c
Pulling from current means we will no longer need to build all
of the packages, instead we would only need to build whatever ZUUL
tells us to build. Because we are no longer building several
packages in the same ci run we can get off our TripleO fork of
delorean as the speed difference will be a lot less.
This commit also includes changes to handle the rename of
ironic-discoverd to ironic-inspector as both sets of changes
are now needed to unblock CI.
Co-Authored-By: John Trowbridge <trown@redhat.com>
Change-Id: Ia879942910c367dc47942f2e0c63351abaf16f63
Depends-on: Id4855d9f4275912440f197f900a59fe5e3bea693
Depends-on: I319f818deeeafc36968b73b8d249c5bbcaa0d2ea