Meeting minutes for 2015-03-30
by Andreas Nilsson
================
#cockpit Meeting
================
Meeting started by andreasn at 13:01:45 UTC. The full logs are available
at
http://meetbot.fedoraproject.org/cockpit/2015-03-30/cockpit.2015-03-30-13...
.
Meeting summary
---------------
* Agenda (andreasn, 13:03:35)
* Fedora 22 (andreasn, 13:06:26)
* GSoC (andreasn, 13:11:52)
* LINK:
https://fedoraproject.org/wiki/Summer_coding_ideas_for_2015#Cockpit_UI_fo...
(sgallagh, 13:14:00)
*
https://fedoraproject.org/wiki/Summer_coding_ideas_for_2015#Cockpit_UI_fo...
(andreasn, 13:14:10)
* Tasks are "Cockpit UI for Rolekit", "Docker Volume Support" and
"Cockpit support for systemd timers" (sgallagh, 13:14:31)
* ACTION: sgallagh and dperpeet to review GSoC applications
(sgallagh, 13:19:03)
* Patternfly migration (andreasn, 13:21:45)
* User syncing without accountsservice (andreasn, 13:24:10)
* storage daemon (andreasn, 13:33:37)
* LINK: https://github.com/cockpit-project/cockpit/pull/2050
(petervo, 13:34:52)
* Open Floor (andreasn, 13:42:10)
Meeting ended at 13:48:51 UTC.
Action Items
------------
* sgallagh and dperpeet to review GSoC applications
Action Items, by person
-----------------------
* dperpeet
* sgallagh and dperpeet to review GSoC applications
* sgallagh
* sgallagh and dperpeet to review GSoC applications
* **UNASSIGNED**
* (none)
People Present (lines said)
---------------------------
* stefw (60)
* andreasn (38)
* sgallagh (23)
* mvollmer (21)
* petervo (13)
* dperpeet (9)
* zodbot (6)
* github (1)
Generated by `MeetBot`_ 0.1.4
.. _`MeetBot`: http://wiki.debian.org/MeetBot
9 years
Weekly meeting, w13
by Marius Vollmer
================
#cockpit Meeting
================
Meeting started by mvollmer at 14:06:41 UTC. The full logs are available
at
http://meetbot.fedoraproject.org/cockpit/2015-03-23/cockpit.2015-03-23-14...
.
Meeting summary
---------------
* Agenda (mvollmer, 14:07:13)
* Fedora 22 Test Day (mvollmer, 14:11:40)
* Fedora bug for Firefox cert handling:
https://bugzilla.redhat.com/show_bug.cgi?id=1204670 (stefw,
14:16:38)
* https://bugzilla.redhat.com/show_bug.cgi?id=1204739 (stefw,
14:20:33)
* ACTION: mvo tell people to go explore (mvollmer, 14:23:23)
* ACTION: mvo phrase firewall opening instructions as a workaround,
with link to bug. (mvollmer, 14:23:45)
* ACTION: mvo check what is in the repos and update page once 0.45 is
available. (mvollmer, 14:24:14)
* Status on removing cockpitd (mvollmer, 14:24:48)
* Suggestion to move to JSON format for list of machines, without
retaining backwards compatibility (stefw, 14:26:47)
* ACTION: Will migrate the machines format file to JSON without
backwards-compatibility ... the plan being that the new format will
be regarded as stable (stefw, 14:30:29)
* Status on javascript bundles (mvollmer, 14:36:48)
* LINK: https://github.com/cockpit-project/cockpit/pull/1980 (stefw,
14:37:47)
* Status of integration tests (mvollmer, 14:47:55)
* LINK: https://github.com/cockpit-project/cockpit/pull/1988 has been
busy for a while (dperpeet, 14:50:06)
* LINK: https://github.com/cockpit-project/cockpit/pull/1988 adds an
env variable to wait a bit (dperpeet, 14:52:20)
* Open Floor (mvollmer, 14:57:07)
* There are applicants to Google Summer of Code who want to work on
Cockpit. (sgallagh, 15:00:05)
* LINK:
https://web.archive.org/web/20140713130957/http://en.flossmanuals.net/GSo...
works as a backup (sgallagh, 15:00:36)
Meeting ended at 15:10:45 UTC.
Action Items
------------
* mvo tell people to go explore
* mvo phrase firewall opening instructions as a workaround, with link to
bug.
* mvo check what is in the repos and update page once 0.45 is available.
* Will migrate the machines format file to JSON without
backwards-compatibility ... the plan being that the new format will be
regarded as stable
Action Items, by person
-----------------------
* **UNASSIGNED**
* mvo tell people to go explore
* mvo phrase firewall opening instructions as a workaround, with link
to bug.
* mvo check what is in the repos and update page once 0.45 is
available.
* Will migrate the machines format file to JSON without
backwards-compatibility ... the plan being that the new format will
be regarded as stable
People Present (lines said)
---------------------------
* stefw (88)
* mvollmer (82)
* dperpeet (30)
* sgallagh (17)
* jscotka (17)
* zodbot (8)
* andreasn_ (6)
* petervo (2)
* github (1)
Generated by `MeetBot`_ 0.1.4
.. _`MeetBot`: http://wiki.debian.org/MeetBot
9 years
Firefox hangs connecting to Cockpit
by Stef Walter
There's a Firefox bug where it hangs for a long time when trying to
connect to Cockpit over TLS.
This is a Firefox issue, and has to do with the TLS certificates that
Cockpit automatically-generates. The same thing happens with Firefox use
of these self-signed certificates on apache or other servers.
Details: https://bugzilla.redhat.com/show_bug.cgi?id=1204670
Cockpit 0.45 has a work-around to generate certificates slightly
differently so that Firefox doesn't fall over.
To regenerate problematic certificates you can:
$ sudo rm /etc/cockpit/ws-certs.d/~self-signed.cert
$ sudo systemctl restart cockpit
Cheers,
Stef
9 years
First signs of the new test infrastructure
by Marius Vollmer
Hi,
on Friday we merged the first piece of the new test infrastructure.
The main goals are:
* Rewrite our tests to work in more environments, such as on bare metal
and as part of existing test efforts. We use Avocado now to define
and run the tests.
* Get rid of our own custom code for defining and running test
machines. We use libvirt now, with virt-deploy.
This is early days still, and we have ported only 4 tests. Those tests
need to pass for a pull request to become "Hubbot green", and I am
watching them closely.
9 years, 1 month
Inheriting avocado.test.Test
by Marius Vollmer
Hi Jan,
as promised, here are some details about the class inheritance trouble I
ran into.
I wanted to have a intermediate class cockpit.CockpitTest that inherits
from avocado.test.Test, and then inherit our actual test classes from
cockpit.CockpitTest.
[ This might or might not be a good idea. In general I agree that
inheritance is a weak tool, but let's pretend for now that I really
need to have this inheritance chain.
]
This is the CockpitTest class:
# cockpit.py
from avocado import test
__all__ = [ 'CockpitTest' ]
class CockpitTest(test.Test):
pass
And this was my first try to use it:
#! /usr/bin/python
# test.py
from avocado import job
from cockpit import CockpitTest
class MyTest(CockpitTest):
def action(self):
self.log.info("Yay!")
if __name__ == "__main__":
job.main()
This leads to this error:
Traceback (most recent call last):
File "/usr/lib/python2.7/site-packages/avocado/test.py", line 332, in action
raise NotImplementedError('Test subclasses must implement an action '
NotImplementedError: Test subclasses must implement an action method
The reason is that Avocado picks cockpit.CockpitTest as the class to
test, not MyTest. I believe this is essentially random: Avocado
enumerates all members of "test" and picks the first that is a subclass
of avocado.test.Test, which happens to be CockpitTest.
This variant of test.py fixes the problem:
#! /usr/bin/python
# test2.py
from avocado import job
import cockpit
class MyTest(cockpit.CockpitTest):
def action(self):
self.log.info("Yay!")
if __name__ == "__main__":
job.main()
Now CockpitTest is no longer a member of the "test" module, and Avocado
will not consider it.
Thus, the test resolution algorithm of Avocado seems to lead to
undefined behavior when there is more than one subclass of
avocado.test.Test.
I think the "one class-under-test per module" rule is fine, but maybe
Avocado could warn/fail when there is more than one?
Also, I didn't find a way to control the test resolution and tell
Avocado to ignore CockpitTest, or explicitly tell it to use MyTest.
The test2.py hack will work for now, but maybe Avocado could be improved
by only considering classes that have been defined in the same module
that it calls getmembers on.
https://github.com/avocado-framework/avocado/pull/474
Thanks!
9 years, 1 month
Meeting Minutes for 2015-03-09
by Andreas Nilsson
================
#cockpit Meeting
================
Meeting started by andreasn_ at 14:02:50 UTC. The full logs are
available at
http://meetbot.fedoraproject.org/cockpit/2015-03-09/cockpit_weekly_meetin...
.
Meeting summary
---------------
* Agenda (andreasn_, 14:04:35)
* Optional PCP update (andreasn_, 14:07:47)
* Avocado updates (andreasn, 14:13:41)
* Subscriptions (andreasn, 14:32:28)
* LINK: https://github.com/cockpit-project/cockpit/pull/1857/
(andreasn, 14:33:38)
* iSCSI initiator (andreasn, 14:37:46)
* LINK:
https://github.com/cockpit-project/cockpit/wiki/Feature:-iSCSI-Initiator
(andreasn, 14:38:20)
* Open Floor (andreasn, 14:41:35)
* LINK: https://fedoraproject.org/wiki/Releases/22/Schedule
(sgallagh, 14:42:40)
* LINK: https://trello.com/c/duRQw1rU/113-milestone-fedora-22
(dperpeet, 14:43:40)
* LINK: http://dl.fedoraproject.org/pub/alt/stage/22_Alpha_RC3/Server/
(sgallagh, 14:51:18)
Meeting ended at 14:54:53 UTC.
Action Items
------------
Action Items, by person
-----------------------
* **UNASSIGNED**
* (none)
People Present (lines said)
---------------------------
* mvollmer (78)
* dperpeet (48)
* andreasn (43)
* sgallagh (17)
* andreasn_ (10)
* petervo (9)
* zodbot (8)
* andreasn__ (3)
* stefw (0)
Generated by `MeetBot`_ 0.1.4
.. _`MeetBot`: http://wiki.debian.org/MeetBot
9 years, 1 month
Cockpit needs removal of packages and reinstall
by Stef Walter
Today, we merged a further simplification of how our Cockpit packages
are loaded:
https://github.com/cockpit-project/cockpit/pull/1869
If you're tracking Cockpit from git master, you'll need to do the following:
$ rm ~/.local/share/cockpit/*
$ sudo rm -rf /usr/share/cockpit/*
$ cd /path/to/cockpit/builddir
$ make
$ sudo make install
Some packages were renamed. So you'll need to remove all references to
the old packages on your system before you continue to use builds from
git master.
If you're tracking Cockpit development via RPMs or other distribution
packages, then this will be handled automatically for you.
Cheers,
Stef
9 years, 1 month
Ubuntu Packaging - "TLS support is not available"
by Jared Sutton
I'm almost ready to upload a usable package based on 0.40 to my PPA
repository, however, I'm getting the following error when trying to start
cockpit...
root@utopic-build:~/cockpit-0.40# systemctl status cockpit -l
cockpit.service - Cockpit Web Server
Loaded: loaded (/lib/systemd/system/cockpit.service; static)
Active: failed (Result: exit-code) since Sun 2015-03-01 15:32:32 EST;
10min ago
Docs: man:cockpit-ws(8)
Process: 21647 ExecStartPre=/usr/sbin/remotectl certificate --ensure
--user=root --group= (code=exited, status=1/FAILURE)
Mar 01 15:32:31 utopic-build remotectl[21647]: Generating temporary
certificate using: openssl req -x509 -days 36500 -newkey rsa:2048 -keyout
/etc/cockpit/ws-certs.d/~self-signed.MRC1UX.tmp -keyform PEM -nodes -out
/etc/cockpit/ws-certs.d/~self-signed.HQC1UX.tmp -outform PEM -subj
/CN=localhost.localdomain
Mar 01 15:32:32 utopic-build remotectl[21647]: remotectl:
/etc/cockpit/ws-certs.d/~self-signed.cert: *TLS support is not available*
Mar 01 15:32:32 utopic-build systemd[1]: cockpit.service: control process
exited, code=exited status=1
Mar 01 15:32:32 utopic-build systemd[1]: Failed to start Cockpit Web Server.
Mar 01 15:32:32 utopic-build systemd[1]: Unit cockpit.service entered
failed state.
I've verified that TLS support is enabled during the configure phase while
building the package, and it's not giving me any pointers as to *why* TLS
support isn't available. Any ideas?
Thanks,
Jared
9 years, 1 month
