Thierry Parmentelat [Fri, 7 Feb 2014 07:41:40 +0000 (08:41 +0100)]
fix qemu-export in dry-run
Thierry Parmentelat [Fri, 7 Feb 2014 07:38:00 +0000 (08:38 +0100)]
decorators maintain __name__ in returned function so that these steps can be ignored
Thierry Parmentelat [Thu, 6 Feb 2014 16:49:57 +0000 (17:49 +0100)]
enfoui is currently down
Thierry Parmentelat [Thu, 6 Feb 2014 15:22:47 +0000 (16:22 +0100)]
re-instate check-vsys-defaults as an ignored step
renamed a few steps for consistency
define macro steps for testing the ability of a qemu box to kick off a node (qemu-again*)
Thierry Parmentelat [Thu, 6 Feb 2014 15:20:52 +0000 (16:20 +0100)]
a sanity check for duplicate vnodes
Thierry Parmentelat [Thu, 6 Feb 2014 11:31:01 +0000 (12:31 +0100)]
nicer msg
Thierry Parmentelat [Thu, 6 Feb 2014 11:30:08 +0000 (12:30 +0100)]
trash Vs-oriented stuff
Thierry Parmentelat [Tue, 4 Feb 2014 09:17:09 +0000 (10:17 +0100)]
renumbered kvm boxes again
Thierry Parmentelat [Fri, 31 Jan 2014 06:27:10 +0000 (07:27 +0100)]
decidedly
Thierry Parmentelat [Thu, 30 Jan 2014 15:57:33 +0000 (16:57 +0100)]
bugfix
Thierry Parmentelat [Thu, 30 Jan 2014 15:16:54 +0000 (16:16 +0100)]
this seems to work for the special f… f14 VMs under f20
I can’t wait to drop support for f14
Thierry Parmentelat [Thu, 30 Jan 2014 14:50:51 +0000 (15:50 +0100)]
cosmetic
Thierry Parmentelat [Thu, 30 Jan 2014 08:02:39 +0000 (09:02 +0100)]
again
Thierry Parmentelat [Thu, 30 Jan 2014 08:01:47 +0000 (09:01 +0100)]
bugfix
Thierry Parmentelat [Wed, 29 Jan 2014 20:55:19 +0000 (21:55 +0100)]
fix quoting issues in the comman-building chain
Thierry Parmentelat [Wed, 29 Jan 2014 20:18:34 +0000 (21:18 +0100)]
get rid of sequels of vs-based boxes
use virsh lxc-enter-namespace instead of ssh to talk to TestPlc
Thierry Parmentelat [Wed, 29 Jan 2014 13:53:56 +0000 (14:53 +0100)]
drop support for vs-hosted plcs
tweak for f14 plcs under f20 hosts
Thierry Parmentelat [Wed, 29 Jan 2014 09:59:37 +0000 (10:59 +0100)]
cosmetic improvement for the daily mail
Thierry Parmentelat [Tue, 28 Jan 2014 12:13:36 +0000 (13:13 +0100)]
renumberd kvm64 boxes
Thierry Parmentelat [Tue, 28 Jan 2014 09:20:43 +0000 (10:20 +0100)]
warhol is now off the hook
Thierry Parmentelat [Tue, 28 Jan 2014 08:53:56 +0000 (09:53 +0100)]
all build boxes run lxc
Thierry Parmentelat [Tue, 28 Jan 2014 07:14:50 +0000 (08:14 +0100)]
put gotan back to business
Thierry Parmentelat [Mon, 27 Jan 2014 13:31:49 +0000 (14:31 +0100)]
print out exit codes
Thierry Parmentelat [Mon, 27 Jan 2014 08:50:00 +0000 (09:50 +0100)]
cleanup
Thierry Parmentelat [Mon, 27 Jan 2014 08:45:24 +0000 (09:45 +0100)]
typo
Thierry Parmentelat [Mon, 27 Jan 2014 08:45:14 +0000 (09:45 +0100)]
manage -t now shows a status line with = B or W (for warning)
Thierry Parmentelat [Mon, 27 Jan 2014 08:26:44 +0000 (09:26 +0100)]
various fixes for the warning mode
Thierry Parmentelat [Sat, 25 Jan 2014 22:49:02 +0000 (23:49 +0100)]
more robust logic for returning retcod when ignored & failure
also revert to the following scheme as some exit(1) were already done elsewhere
# 0: SUCCESS
# 1: FAILURE
# 2: SUCCESS but some ignored steps failed
# 3: OTHER ERROR
Thierry Parmentelat [Fri, 24 Jan 2014 20:09:51 +0000 (21:09 +0100)]
dbg
Thierry Parmentelat [Fri, 24 Jan 2014 17:32:35 +0000 (18:32 +0100)]
implement retcod as adverized
Thierry Parmentelat [Fri, 24 Jan 2014 13:00:53 +0000 (14:00 +0100)]
exit code
# 0: SUCCESS
# 1: SUCCESS but some ignored steps failed
# 2: FAILED STEP
# 3: OTHER ERROR
Thierry Parmentelat [Fri, 24 Jan 2014 10:41:18 +0000 (11:41 +0100)]
one single simpler command to create build or test vms
Thierry Parmentelat [Thu, 23 Jan 2014 19:55:47 +0000 (20:55 +0100)]
no /rootfs/ anymore in test mv path
Thierry Parmentelat [Thu, 23 Jan 2014 11:35:36 +0000 (12:35 +0100)]
typo in comment
Thierry Parmentelat [Thu, 23 Jan 2014 11:35:14 +0000 (12:35 +0100)]
cosmetic
Thierry Parmentelat [Thu, 23 Jan 2014 10:20:03 +0000 (11:20 +0100)]
small fix for ignored steps
Thierry Parmentelat [Thu, 23 Jan 2014 07:43:56 +0000 (08:43 +0100)]
a more thorough implementation for ignored steps
Thierry Parmentelat [Wed, 22 Jan 2014 14:31:12 +0000 (15:31 +0100)]
add ability to define ignorable steps through decorator ignore_result
in default scenario, define check_initscripts and ssh_slice_again as ignored steps so we can try and assess all the rest
Thierry Parmentelat [Tue, 21 Jan 2014 17:29:13 +0000 (18:29 +0100)]
cosmetic
Thierry Parmentelat [Tue, 21 Jan 2014 17:27:50 +0000 (18:27 +0100)]
sensing a lxc build box
Thierry Parmentelat [Tue, 21 Jan 2014 13:58:48 +0000 (14:58 +0100)]
oops bugfix
Thierry Parmentelat [Tue, 21 Jan 2014 12:31:07 +0000 (13:31 +0100)]
calls lbuild-initvm.sh with new signature
Thierry Parmentelat [Fri, 17 Jan 2014 14:05:29 +0000 (15:05 +0100)]
add a test that /vservers is not found in slice - this happens right now, as slices do not run in the right mount namespace (at least not when reached through ssh)
Thierry Parmentelat [Thu, 16 Jan 2014 19:45:55 +0000 (20:45 +0100)]
oops
Thierry Parmentelat [Thu, 16 Jan 2014 19:45:13 +0000 (20:45 +0100)]
Merge branch 'sfa-geni-v3' of ssh://git.onelab.eu/git/tests into sfa-geni-v3
Thierry Parmentelat [Thu, 16 Jan 2014 19:44:58 +0000 (20:44 +0100)]
gitignore
Mohamed Larabi [Thu, 16 Jan 2014 17:21:26 +0000 (18:21 +0100)]
remove gotan from test infra cause tests are passing yet on it
Mohamed Larabi [Thu, 16 Jan 2014 15:53:21 +0000 (16:53 +0100)]
put bacl deathvegas
Mohamed Larabi [Thu, 16 Jan 2014 14:11:44 +0000 (15:11 +0100)]
put gotan as lxc test box and remove deathvegas
Thierry Parmentelat [Tue, 14 Jan 2014 13:30:04 +0000 (14:30 +0100)]
remove dbg msg
Thierry Parmentelat [Sun, 12 Jan 2014 23:01:46 +0000 (00:01 +0100)]
re-check slice creation once they have been torn down to make sure a second run with the same name is fine
Thierry Parmentelat [Fri, 10 Jan 2014 09:16:34 +0000 (10:16 +0100)]
fix changes to the build for vs/lxc
Thierry Parmentelat [Thu, 9 Jan 2014 16:59:43 +0000 (17:59 +0100)]
cosmetic
Thierry Parmentelat [Thu, 9 Jan 2014 16:58:00 +0000 (17:58 +0100)]
specify local substrate with vs-based and lxc-based build boxes separately
Thierry Parmentelat [Thu, 9 Jan 2014 16:25:56 +0000 (17:25 +0100)]
use new names for initvm creation scripts
Thierry Parmentelat [Fri, 20 Dec 2013 13:37:22 +0000 (14:37 +0100)]
turn off kruder that is down too (hdd failure)
Thierry Parmentelat [Fri, 20 Dec 2013 10:08:34 +0000 (11:08 +0100)]
turn off gotan for now
Thierry Parmentelat [Wed, 11 Dec 2013 10:25:52 +0000 (11:25 +0100)]
store result of initscript in /var/tmp that is not in memory, for easier troubleshooting
change various names for more clarity
Thierry Parmentelat [Wed, 11 Dec 2013 10:08:15 +0000 (11:08 +0100)]
create a config_no_omf wrapper so we can drop the no_omf branch
Thierry Parmentelat [Wed, 11 Dec 2013 09:56:26 +0000 (10:56 +0100)]
add timestamps when running initscripts
Thierry Parmentelat [Tue, 10 Dec 2013 09:09:34 +0000 (10:09 +0100)]
Setting tag tests-5.3-1
this is for testing myplc/plcapi 5.3.x with sfa-v3
Thierry Parmentelat [Tue, 10 Dec 2013 09:08:26 +0000 (10:08 +0100)]
bump version to 5.3
Mohamed Larabi [Tue, 12 Nov 2013 15:55:27 +0000 (16:55 +0100)]
Set PLC_HRN_ROOT in plc config
Thierry Parmentelat [Tue, 27 Aug 2013 12:48:58 +0000 (14:48 +0200)]
Merge branch 'master' into sfa-geni-v3
Thierry Parmentelat [Tue, 27 Aug 2013 11:04:56 +0000 (13:04 +0200)]
get ssh-slice and ssh-slice-off to work in parallel on all 3 slices
Thierry Parmentelat [Tue, 27 Aug 2013 11:02:04 +0000 (13:02 +0200)]
another cosmetic
Thierry Parmentelat [Tue, 27 Aug 2013 09:52:22 +0000 (11:52 +0200)]
cosmetic, fix message when ssh_slice_off fails
Thierry Parmentelat [Mon, 26 Aug 2013 10:43:52 +0000 (12:43 +0200)]
cosmetic changes to manage's output wrt one-line summary of lxc and qemu boxes - shows uname() only with -v
Thierry Parmentelat [Sat, 13 Jul 2013 22:16:37 +0000 (00:16 +0200)]
Setting tag tests-5.2-7
only inria's local settings
Thierry Parmentelat [Wed, 10 Jul 2013 07:15:01 +0000 (09:15 +0200)]
10 VM's per plc box should be enough and hopefully bring the getty-related load a bit down
Thierry Parmentelat [Wed, 26 Jun 2013 07:22:37 +0000 (09:22 +0200)]
Setting tag tests-5.2-6
substrate sensing more robust to lxc containers with no timestamp
Thierry Parmentelat [Tue, 25 Jun 2013 07:17:21 +0000 (09:17 +0200)]
Merge branch 'master' into sfa-geni-v3
Thierry Parmentelat [Tue, 25 Jun 2013 06:49:16 +0000 (08:49 +0200)]
be robust against running lxc containers without a timestamp
Thierry Parmentelat [Tue, 11 Jun 2013 09:47:42 +0000 (11:47 +0200)]
Setting tag tests-5.2-5
add a third slice per plc that is omf-friendly
currently this ends up broken after the tests (but undetected)
because nodemanager needs a fix to unmount ssh keys exposed in the slice
(the tests do unpopulate / repolupate the slices with empty/fill_slices)
Thierry Parmentelat [Thu, 6 Jun 2013 08:36:29 +0000 (10:36 +0200)]
each plc to create a third, omf-friendly slice
Thierry Parmentelat [Tue, 4 Jun 2013 09:47:58 +0000 (11:47 +0200)]
Setting tag tests-5.2-4
more robust and more efficient substrate sensing
manage -tv now detects very early failures instead of reporting success
mod_python step defined but not enabled
Thierry Parmentelat [Mon, 3 Jun 2013 10:48:06 +0000 (12:48 +0200)]
mod_python is broken for now
Thierry Parmentelat [Fri, 31 May 2013 22:24:19 +0000 (00:24 +0200)]
mod_python is broken for now
Thierry Parmentelat [Fri, 31 May 2013 13:17:02 +0000 (15:17 +0200)]
Merge branch 'master' into sfa-geni-v3
Thierry Parmentelat [Fri, 31 May 2013 12:26:04 +0000 (14:26 +0200)]
insert step mod_python so we stay away from broken wsgi
Thierry Parmentelat [Thu, 30 May 2013 10:56:08 +0000 (12:56 +0200)]
improve handling of unreachable boxes
Thierry Parmentelat [Thu, 30 May 2013 10:47:28 +0000 (12:47 +0200)]
show load for qemu boxes
Thierry Parmentelat [Thu, 30 May 2013 10:42:14 +0000 (12:42 +0200)]
manage more efficient (probe all in once) and with nicer output
Thierry Parmentelat [Wed, 29 May 2013 10:34:02 +0000 (12:34 +0200)]
fix testmaster sensing : tests that break very early (like, when the IP pool is full) were reported as successful, which of course is very wrong
Thierry Parmentelat [Wed, 29 May 2013 09:42:48 +0000 (11:42 +0200)]
no need for a vs-box for testing plcs any more
Mohamed Larabi [Tue, 14 May 2013 17:16:38 +0000 (19:16 +0200)]
Reduce max number of PLC's on LXCboxes
Thierry Parmentelat [Thu, 2 May 2013 08:08:11 +0000 (10:08 +0200)]
Setting tag tests-5.2-3
new separate step qemu_clean_mine
default is to use an lxc-based vplc box (options -S and -X)
minor changes
Thierry Parmentelat [Thu, 25 Apr 2013 12:30:42 +0000 (14:30 +0200)]
Merge branch 'master' into sfa-geni-v3
Thierry Parmentelat [Wed, 24 Apr 2013 14:46:47 +0000 (16:46 +0200)]
Merge branch 'master' of ssh://git.onelab.eu/git/tests
Thierry Parmentelat [Wed, 24 Apr 2013 14:46:35 +0000 (16:46 +0200)]
has option -X (old) and -S (new) to select between a
vs- or an lxc-hosted vplc - default is lxc
Thierry Parmentelat [Wed, 24 Apr 2013 14:45:56 +0000 (16:45 +0200)]
split qemu_export into clean & export, make it safer to re-use run_log
Mohamed Larabi [Wed, 24 Apr 2013 12:40:39 +0000 (14:40 +0200)]
put back deathvegas
Thierry Parmentelat [Wed, 24 Apr 2013 12:23:03 +0000 (14:23 +0200)]
turn deathvegas back off
Thierry Parmentelat [Wed, 24 Apr 2013 08:30:03 +0000 (10:30 +0200)]
deathvegas reinstalled
Thierry Parmentelat [Tue, 23 Apr 2013 14:47:51 +0000 (16:47 +0200)]
Setting tag tests-5.2-2
kvm boxes now run f18 and use /vservers/ as their big disk area like all other boxes
lxc boxes run 1.0.4, which required a change in plc.d/gpg - only gotan for now
Thierry Parmentelat [Mon, 22 Apr 2013 14:20:04 +0000 (16:20 +0200)]
try libvirt1.0.4 now that /dev/random is managed differently
Thierry Parmentelat [Fri, 19 Apr 2013 07:39:30 +0000 (09:39 +0200)]
kvm1 back in business
Thierry Parmentelat [Thu, 18 Apr 2013 19:16:49 +0000 (21:16 +0200)]
kvm64-1 has its qemu nodes hang on partitioning disk...
Thierry Parmentelat [Thu, 18 Apr 2013 09:51:24 +0000 (11:51 +0200)]
Merge branch 'master' into sfa-geni-v3
Thierry Parmentelat [Thu, 18 Apr 2013 09:51:24 +0000 (11:51 +0200)]
Merge branches 'master' and 'master' of ssh://git.onelab.eu/git/tests