[29-02-2016] Weekly meeting minutes
by David Caro
Discussions (continued):
========================
* oVirt sdk version
* On oVirt 4.0 two versions of the sdk will be supported, 3 and 4, api
version 3 is provided by the ovirt-sdk-3.6 rpms, while the 4 is provided
by the 4.0 ones (not ready yet), so we will be using the 3.6 until 4.0
ones are ready, when we will start to adapt to it with oVirt 5.0 as
deadline
* Change from json to yaml
* Maybe support both at first
* Nicer multiline
* Comments
* Nicer quoting
**Approved**
* Userland runs (running lago without special permissions, qemu:///session)
* Network is an issue
* http://blog.wikichoon.com/2016/01/qemusystem-vs-qemusession.html
* https://github.com/pradels/vagrant-libvirt/issues/272
* http://comments.gmane.org/gmane.comp.emulators.libvirt/117725
* ^ have to chat with the libvirt guys on how to use that
* You have to create the bridge nets with superuser privs, and then
use the bridges from the conf:
* A dump of the domain xml shows:
<interface type='bridge'>
<mac address='52:54:00:91:98:c6'/>
<source bridge='virbr0'/>
<target dev='tap0'/>
<model type='virtio'/>
<alias name='net0'/>
<address type='pci' domain='0x0000' bus='0x00' slot='0x03' function='0x0'/>
</interface>
* There’s a default bridge (virbr0) always created and available on fedora
boxes (by ways of /etc/qemu/bridge.conf), is what gnome-boxes uses
* Might be nice to have, even requesting root pass to create bridges when
needed
* Multiple ‘envs’ in the same prefix workflows, now that .lago is the default
prefix path (see
* Should we use multiple envs per prefix? or one prefix per env?
* Allow selecting the environment, and the prefix dir
**Approved**
* layered images for vms (will probably have to develop ourselves, at least,
the client side)
* Make it work on el7 too (maybe just open/push bzs to bring in the needed
libguestfs features
* Export/import of vm images from prefix? On the long run, being able to
export/import full prefixes?
* Very interesting to have
* Ovirt:
* ovirt-standard-ci plugin to build vdsm/ovirt rpms
* Being able to run/deploy live vdsm/ovirt (unpackaged)
* Drop reposync for now
* Run engine in a container/chroot + vdsm in a vm?
* Action items:
* Start a thread to discuss the reposync drop
* Open github issues for all the agreed features
* Change the ovirt 4.0 publisher to publish the 3.6 sdk (until 4.0 is
ready, then publish both on 4.0 repo)
--
David Caro
Red Hat S.L.
Continuous Integration Engineer - EMEA ENG Virtualization R&D
Tel.: +420 532 294 605
Email: dcaro(a)redhat.com
IRC: dcaro|dcaroest@{freenode|oftc|redhat}
Web: www.redhat.com
RHT Global #: 82-62605
8 years, 8 months
3.6 suddenly fails
by Yaniv Kaul
Any idea what could be the problem here:
@ Deploy oVirt environment:
# [Thread-2] Deploy VM lago_basic_suite_3_6_host1:
# [Thread-3] Deploy VM lago_basic_suite_3_6_engine:
# [Thread-4] Deploy VM lago_basic_suite_3_6_storage-iscsi:
# [Thread-5] Deploy VM lago_basic_suite_3_6_storage-nfs:
* [Thread-2] Wait for ssh connectivity:
# [Thread-6] Deploy VM lago_basic_suite_3_6_host0:
* [Thread-3] Wait for ssh connectivity:
* [Thread-4] Wait for ssh connectivity:
* [Thread-5] Wait for ssh connectivity:
* [Thread-6] Wait for ssh connectivity:
* [Thread-5] Wait for ssh connectivity: Success (in 0:01:15)
* [Thread-5] Run script setup_storage_nfs.sh:
* [Thread-4] Wait for ssh connectivity: Success (in 0:01:17)
* [Thread-4] Run script setup_storage_iscsi.sh:
* [Thread-5] Run script setup_storage_nfs.sh: Success (in 0:00:09)
* STDERR
mke2fs 1.42.8 (20-Jun-2013)
mount: wrong fs type, bad option, bad superblock on /dev/vdc1,
missing codepage or helper program, or other error
In some cases useful info is found in syslog - try
dmesg | tail or so.
# [Thread-5] Deploy VM lago_basic_suite_3_6_storage-nfs: ERROR (in
0:01:25)
Error while running thread
Traceback (most recent call last):
File "/usr/lib/python2.7/site-packages/lago/utils.py", line 45, in
_ret_via_queue
queue.put({'return': func()})
File "/usr/lib/python2.7/site-packages/ovirtlago/__init__.py", line 469,
in _deploy_host
host.name(),
RuntimeError:
/home/mini/ovirt-system-tests/basic_suite_3.6/deploy-scripts/setup_storage_nfs.sh
failed with status 64 on lago_basic_suite_3_6_storage-nfs
8 years, 9 months
[26-02-2016] Weekly lago release - 0.10.0
by David Caro
Hi everyone!
This week we have a couple of very interesting new features:
* Using .lago as default prefix for init, that means that if you run init
with just the json configuration file, it will create the prefix at
$PWD/.lago, but, if you did so, then any command that you run on any dir
that's under the $PWD dir that has the .lago prefix, will be run on that
prefix!
For example:
> cd /my/awesome/project
> lago init myenv.json
# this create the prefix at /my/awesome/project/.lago
> cd src/mymodule
> lago status
# this will run on the /my/awesome/project/.lago prefix automatically!
> lago start
> lago shell myel7_vm
...
* Ouptut format for info gathering commands, for now, just the status and
snapshot -l commands. Now you can add the --out-format|-f parameter to
lago, and when running a command to extract info from the prefix, it will
format the ouptut (yaml, json and 'default' are implemented for now). For
example:
> lago -f json status
{
"Prefix": {
"Base directory": ...,
"Networks": {
"n0": {
"gateway": "192.168.201.1",
"management": true,
"status": "down"
},
...
},
"UUID": "fe689d3c39d011e5b22b54ee755a00ca",
"VMs": {
"vm0": {
"NICs": {
...
},
"VNC port": null,
"distro": "cirros",
"metadata": {},
"root password": "123456",
"snapshots": "",
"status": "down"
}
}
}
}
* Added the public template server as the default template repo, so you don't
have to supply it each time
* Using 'direct' as default libguestfs backend, that should get rid of the
annoying errors when bootstrapping the domains on prefix initialization.
The full changelog:
http://resources.ovirt.org/repos/lago/stable/0.0/CHANGELOG_0.0.txt
--
David Caro
Red Hat S.L.
Continuous Integration Engineer - EMEA ENG Virtualization R&D
Tel.: +420 532 294 605
Email: dcaro(a)redhat.com
IRC: dcaro|dcaroest@{freenode|oftc|redhat}
Web: www.redhat.com
RHT Global #: 82-62605
8 years, 9 months
Fwd: [19-02-2016] Weekly lago release - 0.9.0
by David Caro
Hi everyone!
This week we have a couple of very interesting new features:
* Using .lago as default prefix for init, that means that if you run init
with just the json configuration file, it will create the prefix at
$PWD/.lago, but, if you did so, then any command that you run on any dir
that's under the $PWD dir that has the .lago prefix, will be run on that
prefix!
For example:
> cd /my/awesome/project
> lago init myenv.json
# this create the prefix at /my/awesome/project/.lago
> cd src/mymodule
> lago status
# this will run on the /my/awesome/project/.lago prefix automatically!
> lago start
> lago shell myel7_vm
...
* Ouptut format for info gathering commands, for now, just the status and
snapshot -l commands. Now you can add the --out-format|-f parameter to
lago, and when running a command to extract info from the prefix, it will
format the ouptut (yaml, json and 'default' are implemented for now). For
example:
> lago -f json status
{
"Prefix": {
"Base directory": ...,
"Networks": {
"n0": {
"gateway": "192.168.201.1",
"management": true,
"status": "down"
},
...
},
"UUID": "fe689d3c39d011e5b22b54ee755a00ca",
"VMs": {
"vm0": {
"NICs": {
...
},
"VNC port": null,
"distro": "cirros",
"metadata": {},
"root password": "123456",
"snapshots": "",
"status": "down"
}
}
}
}
* Added the public template server as the default template repo, so you don't
have to supply it each time
* Using 'direct' as default libguestfs backend, that should get rid of the
annoying errors when bootstrapping the domains on prefix initialization.
The full changelog:
http://resources.ovirt.org/repos/lago/stable/0.0/CHANGELOG_0.0.txt
--
David Caro
Red Hat S.L.
Continuous Integration Engineer - EMEA ENG Virtualization R&D
Tel.: +420 532 294 605
Email: dcaro(a)redhat.com
IRC: dcaro|dcaroest@{freenode|oftc|redhat}
Web: www.redhat.com
RHT Global #: 82-62605
8 years, 9 months
Testing oVirt 4.0 with Lago - which SDK should we use?
by Yaniv Kaul
Currently, the SDK that Lago uses is (at least on my Fedora):
ovirt-engine-sdk-python-3.5.1.0-2.fc23.noarch
I think we may need to bring in a newer SDK.
1. Which one?
2. If master - is it backward compatible to run against 3.x already?
3. Can we have them side by side perhaps?
TIA,
Y.
8 years, 9 months
[ovirt-system-tests] What is the correct value of the engine's answer file entry OVESETUP_CONFIG/fqdn ?
by Yaniv Kaul
[ykaul@ykaul answer-files]$ pwd
/home/ykaul/ovirt-system-tests/common/answer-files
[ykaul@ykaul answer-files]$ grep -R "OVESETUP_CONFIG/fqdn" * |sort
el6_3.5.conf:OVESETUP_CONFIG/fqdn=str:engine
el6_3.6.conf:OVESETUP_CONFIG/fqdn=str:*localhost.localdomain*
el6_master.conf:OVESETUP_CONFIG/fqdn=str:engine
el7_3.5.conf:OVESETUP_CONFIG/fqdn=str:engine
el7_master.conf:OVESETUP_CONFIG/fqdn=str:engine
Is the value localhost.localdomain correct for el6-3.6 (only?).
TIA,
Y.
8 years, 9 months
[19-02-2016] Weekly lago release - 0.9.0
by David Caro
Hi everyone!
This week release is small, and has not big changes, but a big bump was done on
adding some functional tests and starting to move toward py.test for unit
tests, the relevant changes:
dcaf87c6: Moving to py.test
840d3956: Fix start a bridged network that's already alive
a20cff51: Avoiding ssh connection for hosts that have no ovirt-scripts
gned
bbcebcd0: Adding support for ovirt node specific VM's
You can get it fresh from:
http://resources.ovirt.org/repos/lago/stable/0.0/rpm/
The full changelog:
http://resources.ovirt.org/repos/lago/stable/0.0/CHANGELOG_0.0.txt
Enjoy!
--
David Caro
Red Hat S.L.
Continuous Integration Engineer - EMEA ENG Virtualization R&D
Tel.: +420 532 294 605
Email: dcaro(a)redhat.com
IRC: dcaro|dcaroest@{freenode|oftc|redhat}
Web: www.redhat.com
RHT Global #: 82-62605
8 years, 9 months
Re: [lago-devel] Fwd: test failed, Failed to connect to the engine
by Yaniv Kaul
On Thu, Jan 28, 2016 at 3:01 PM, Yaniv Kaul <ykaul(a)redhat.com> wrote:
>
>
> On Thu, Jan 28, 2016 at 2:56 PM, Maor Lipchuk <mlipchuk(a)redhat.com> wrote:
>
>>
>>
>>
>>
>> ----- Original Message -----
>> > From: "Yaniv Kaul" <ykaul(a)redhat.com>
>> > To: "Maor Lipchuk" <mlipchuk(a)redhat.com>
>> > Cc: lago-devel(a)ovirt.org
>> > Sent: Thursday, January 28, 2016 12:59:38 PM
>> > Subject: Re: [lago-devel] Fwd: test failed, Failed to connect to the
>> engine
>> >
>> > On Thu, Jan 28, 2016 at 12:30 PM, Maor Lipchuk <mlipchuk(a)redhat.com>
>> wrote:
>> >
>> > > Hi all,
>> > >
>> > > I've been trying to use the LAGO setup, all worked well up to the
>> point
>> > > which I got an error on one of the tests :
>> > >
>> > > Run test: 002_bootstrap.py: ERROR
>> > > Error occured, aborting
>> > > ....
>> > > ... line 82, in do_ovirt_runtest
>> > > raise RuntimeError('Some tests failed')
>> > > RuntimeError: Some tests failed
>> > >
>> > > Looking in the nosetests it looks like it failed to connect to the
>> engine:
>> > >
>> > > File "usr/lib/python2.7/site-packaged/pvortlago/virt.py", line 106,
>> in
>> > > _get_api
>> > > raise RuntimeError('Failed to connect to the engine')
>> > >
>> >
>> > Can you check if the engine is up? In my case, it was not installed, as
>> the
>> > answer file was incorrect.
>> > See https://bugzilla.redhat.com/show_bug.cgi?id=1302045 for details and
>> > solution.
>> > Y.
>>
>> I don't see that ovirt-engine is installed on my env, but I do see that
>> the ovirt-release36-002-2 is installed.
>> I was trying to set the boolean to false as was mentioned in the bug to
>> see if that make any difference but still didn't had much luck.
>>
>
> Need to look at the engine's setup log, then.
>
I prepared a patch to the installation test that:
1. Verifies the installation succeeded (based on error code from the
installation)
2. Verifies the ovirt-engine service is up.
Only if both succeed, the test passes.
I've cloned the specific bug above to a general bug @
https://bugzilla.redhat.com/show_bug.cgi?id=1303973
Y.
> Y.
>
>
>>
>> >
>> > >
>> > > tests="6" error="1"
>> > >
>> > > Is it only a test issue or is there something wrong with my engine?
>> > >
>> > > Regards,
>> > > Maor
>> > >
>> > >
>> > > _______________________________________________
>> > > lago-devel mailing list
>> > > lago-devel(a)ovirt.org
>> > > http://lists.ovirt.org/mailman/listinfo/lago-devel
>> > >
>> >
>>
>
>
8 years, 9 months
Lago mailing list archives
by David Caro
Hi everyone!!
Last week we realized that the archives for the mailing list were not
working, and that has been solved today. Painfully all the emails that
were sent before were not actually archived so they have been lost in
the archives, but from now on they will be properly archived and
backed up.
Sorry for the inconvenience :/
--
David Caro
Red Hat S.L.
Continuous Integration Engineer - EMEA ENG Virtualization R&D
Tel.: +420 532 294 605
Email: dcaro(a)redhat.com
IRC: dcaro|dcaroest@{freenode|oftc|redhat}
Web: www.redhat.com
RHT Global #: 82-62605
8 years, 9 months