A lot of libvirt interactions with QEMU are via the QEMU monitor
console, which allows you to either query or modify the state of a
virtual machine. Spefici examples include: querying the status of live
block operations, live snapshot operations, live migration, etc.
Enabling the 'qemu_monitor' log filter allows us to capture precisely
what commands libvirt is sending to QEMU.
Note that the log level was intentionally set to '1' (i.e. debug) for
this specific filter, because (a) it's not extremely verbose, (b) when
something breaks, it's helpful to have the exact sequence of
interactions between libvirt and QEMU.
Change-Id: Iba95b6bd7c9f197c8d48c7d978f538e50d4e31fa
There are a couple of targets in here that some might find useful in
doing DevStack testing in multiple remote VMs.
There are some of the usual boring targets too, like stack and unstack,
that do exactly what you would expect.
Change-Id: I7974cac4cc527bacf6f183ac1f344428b05f2fdc
This previously defatuled to 'private' and aligned with devstack's
defaults but it has since been updated to 'None'. This sets the config
value according to devstack's.
Change-Id: I3f480d5480521a93992bedfe602eb20a4999263d
Closes-bug: #1438415
This is needed otherwise we fall back to the nbd code path which is
super hacky. This shouldn't have been deleted for the venv path.
Change-Id: If5cb6cb4944bd0ed3548d53c98443b76725d1c0c
PyEClib was introduced recently to swift-master. It tries
to build liberasurecode which requires the `make` binary.
Change-Id: I8acfed4f7b46a29eac36f6acbe1d66e7fff800db
Reading through the docs for the first time, the reader encounters an
instruction to provide a minimal configuration, with a link that they'd
expect to tell them how to do this.
At present the link actually takes them to the top of
configuration.html, where they read some history about how devstack's
configuration has changed over time.
This is interesting and important and should be in the docs - but in my
opinion a link about setting up a minimal configuration would be more
useful if it takes me to a place that tells them about a minimal
configuration.
To get this, I've had to an an explicit link target into
configuration.rst. I'm not hugely keen on this approach, as I don't
think it scales well. I'd be open to suggestions about other
approaches. The only idea I've had so far though is to simply move the
minimal configuration section right to the top of the page, so that a
link to the doc is a link to the minimal config - the historical
information could be moved to its own topic somewhere further down the doc.
Change-Id: I231ca1b7f17b55f09a4e058dab8ee433893f737e
The public_endpoint and admin_endpoint options are used to set the base
hostname when listing versions. If they are not provided then the
response will use the same hostname as the request was made with - which
is almost always what you actually want.
This means devstack will respond correctly to a version list when behind
a floating IP for example.
Change-Id: Idc48b9d7bee9751deb24d730fdc560b163f39dfe
Rootwrap shouldn't be a unique snowflake. Plus the binaries tend
to be called assuming PATH will find them. Not so with venvs
so we need to work around that brokenness.
Configure Cinder and Nova to use configure_rootwrap().
Change-Id: I8ee1f66014875caf20a2d14ff6ef3672673ba85a
pyOpenSSL has done a rewrite of itself in Python. This may be good
for many reasons, but memory usage is not one of them. It just about
doubles the size of at least swift, which usually consumes about 6% of
a CI testing vm's 8gb RAM. This is enough to push centos hosts into
OOM conditions and then everything falls apart.
The distro packages of pyOpenSSL are the older C-based versions, which
doesn't bring in the kitchen sink of cffi & pycparser.
Change-Id: Icd4100da1d5cbdb82017da046b00b9397813c2f2
A recently merged change Ie35cb537bb670c4773598b8db29877fb8a12ff50
and I71e2594288bae1a71dc2c8c3fb350b913dbd5e2c broke Q_USE_PUBLIC_VETH.
This commit fixes the regression.
Closes-Bug: #1436637
Change-Id: I1447bf98607143ba4954ce5ec3ed94010320baa5
Install a couple of optional feature prereqs in hypervisor plugins.
rootwrap is horribly called indirectly via PATH. The choice, other than fixing
such nonsense, is to force the path in sudo.
TODO:
* python guestfs isn't in pypi, need to specifically install it to not
use the distro package
Change-Id: Iad9a66d8a937fd0b0d1874005588c702e3d75e04
The Linux-IO is a modern way of handling targets.
Per the IRC discussions lioadm as default
seams like a better default for everyone, for now it will be
optional, but the tgtadm admin support expected to be removed when
lioadm works well with all CI (including third party).
Change-Id: Ia54c59914c1d3ff2ef5f00ecf819426bc448d0a9
during the glusterfs integration it was seen that plugins might need
to set new defaults on projects before the project files load. Create
a new override-defaults phase for that.
Intentionally not adding to the documentation yet until we're sure
this works right in the glusterfs case.
Reported-By: Deepak C Shetty <deepakcs@redhat.com>
Change-Id: I13c961b19bdcc1a99e9a7068fe91bbaac787e948
Make it possible for someone to config
PIP_UPGRADE=True
in local.conf and thus force pip_install calls to upgrade. In
automated testing this is probably a bad idea, but in manual testing
or situations where devstack is being used to spin up proof of
concepts having the option to use the latest and greatest Python
modules is a useful way of exploring the health of the ecosystem.
To help with visibility of the setting, and section has been added
in configuration.rst near other similar settings.
Change-Id: I484c954f1e1f05ed02c0b08e8e4a9c18558c05ef
When running Neutron on a single node that only has a single interface,
the following operations are required:
* Remove the IP address from the physical interface
* Add the interface to the OVS physical bridge
* Add the IP address from the physical interface to the OVS bridge
* Update the routing table
The reverse is done on cleanup.
In order run Neutron on a single interface, the $PUBLIC_INTERFACE and
$OVS_PHYSICAL_BRIDGE variables must be set.
Co-Authored-By: Brian Haley <brian.haley@hp.com>
Change-Id: I71e2594288bae1a71dc2c8c3fb350b913dbd5e2c
This commit removes the compute-admin section from the tempest config
file that devstack generates. These options have been removed from
the tempest config and aren't being used, so there is no reason to
keep them around.
Change-Id: I7500fe3b329b913c60fa505a5230db4a5d35d7f1
isset function was moved to config file related functions by accident,
this change also simplfies the isset in a bash >=4.2 way.
All supported distro has at least bash 4.2. (RHEL6 used 4.1)
Change-Id: Id644b46ff9cdbe18cde46e96aa72764e1c8653ac
This commit adds a new flag, TEMPEST_HAS_ADMIN, to enable or disable
setting admin credentials in the tempest config file. This allows for
devstack / tempest configurations where tempest doesn't have admin to
ensure it would work in public cloud scenarios.
Change-Id: Id983417801e4b276334fb9e700f2c8e6ab78f9ba
Adding `uname -a` to stack.sh to make easy to see from the devstack logs
what was the actually running kernel version.
Change-Id: I0068504bf055a588b155b0a60215440d365bf53e
