Commit Graph

35279 Commits

Author SHA1 Message Date
Kevin Fenzi
4f19774152 clevis is fixed, drop these variables that are no longer needed.
Signed-off-by: Kevin Fenzi <kevin@scrye.com>
2021-07-13 15:05:14 -07:00
Peter Buffon
2e41dc05a8 Batch update of bvmhost* Ansible host_vars files for linux-system-roles/network conversion. 2021-07-13 21:57:09 +00:00
Kevin Fenzi
f4e7c25513 clevis: use dns1 instead of dns on buildhw-x86
Signed-off-by: Kevin Fenzi <kevin@scrye.com>
2021-07-13 14:53:53 -07:00
Kevin Fenzi
15860c4950 clevis: fix buildhw-x86 clevis more.
We need to use eth0_ipv4 versions of these variables when eth0_ipv4 is
set.

Signed-off-by: Kevin Fenzi <kevin@scrye.com>
2021-07-13 14:51:34 -07:00
Brendan Early
b1810a1525 Add missing images for fedmsg 2021-07-13 19:52:24 +00:00
Adrian Reber
a0d81cdb4d Add codecs repository definition to scan-primary-mirror
Signed-off-by: Adrian Reber <adrian@lisas.de>
2021-07-13 18:08:00 +02:00
Adrian Reber
382189e844 Do not exclude 'source' when scanning the primary mirror
Signed-off-by: Adrian Reber <adrian@lisas.de>
2021-07-13 18:07:51 +02:00
Pavel Raiskup
fddd881561 copr-be-dev: newer ppc64le image with faster network convergence? 2021-07-13 14:29:32 +02:00
Pavel Raiskup
5f187f46f5 copr-hv: update README for the required grub.cfg update 2021-07-13 14:13:22 +02:00
Pavel Raiskup
2f1b409600 copr-be: ppc64le builders to have hcn-service.init disabled
NetworkManager-wait-online takes too much time, and fails eventually
anyway.
2021-07-13 13:38:26 +02:00
Pavel Raiskup
c7c983063a copr-be: try to ping6 to wake up ipv6 stack early
Seems like some ppc64le boxes start and seem fine, but they are not
accessible over ipv6 for a long time so the spawner timeouts.  When I
tried to manually ping6 from the box outside - the connection from the
outside started to work immediately - so give it an automatic try.
2021-07-13 12:30:27 +02:00
Pavel Raiskup
0bc8c0a9c1 copr-be: sync wait-for-ssh with upstream 2021-07-13 11:16:44 +02:00
Pavel Raiskup
62b824d4a7 copr-be: correct the ppc64le image reference 2021-07-13 10:43:21 +02:00
Pavel Raiskup
8aaa06b0dc copr-be: typo in config 2021-07-13 10:08:51 +02:00
Pavel Raiskup
e03410cb97 copr-be: turn on ppc64le in production 2021-07-13 09:15:50 +02:00
Kevin Fenzi
738c89d04d vmhost-p08-copr02: add second power8 box in rdu-cc
Signed-off-by: Kevin Fenzi <kevin@scrye.com>
2021-07-12 19:07:52 -07:00
lrossett
ee835b9b22 prod and stg vars 2021-07-12 19:54:38 +00:00
lrossett
2d209bef3a resultsdb-ci-listener playbook and role 2021-07-12 19:54:38 +00:00
Pavel Raiskup
56333c899f copr-be: one more ppc64le image update
This one should have enough space in / directory.
2021-07-12 18:10:16 +02:00
Pavel Raiskup
9b28b08059 copr-be: fix ppc64le VM termination 2021-07-12 18:06:07 +02:00
Pavel Raiskup
8903ac9c3f copr-be: disable @copr/copr on devel builders for now
THe @copr/copr doesn't have ppc64le chroots available, so we can not
even enable them.
2021-07-12 17:21:32 +02:00
Pavel Raiskup
d081f58518 copr-be: use copr ansible plugin for enabling @copr/copr 2021-07-12 17:16:13 +02:00
Pavel Raiskup
b94611c483 copr-be: fixed ppc64le image for ipv6 config, #1 2021-07-12 16:23:31 +02:00
Pavel Raiskup
31b5478679 copr-be: don't fail enable-swap when not necessary
Ppc64le images have tsflags=nodocs set, so "rpm --setperms" fails with

    chmod: cannot access '/usr/share/man/man1/copr-rpmbuild.1.gz': No such file or directory

Also, there's a new directory ../workspace, see:
52e8393ce4
2021-07-12 14:37:20 +02:00
Pavel Raiskup
5ab3f72439 copr-hv: more verbose libvirt-new allocator 2021-07-12 14:26:40 +02:00
Pavel Raiskup
96629a900d copr-hv: fix VMs IPv6 allocation
The bug was that we tried to allocate addresses like "*cafe:c1001" (one
additional number).  This is now fixed, and since we are allowed to use
c1, c2, c3, .. let's give each hypervisor a fixed block of 64 addresses,
and don't be afraid to overflow there with new hypervisors.
2021-07-12 14:16:29 +02:00
Mark O Brien
a37980bb0b make dir for sysadmin-openshift group 2021-07-12 10:47:40 +01:00
Pavel Raiskup
69f856149e copr-be-dev: try to start ppc64le machines 2021-07-12 10:22:48 +02:00
Pavel Raiskup
e931ead047 copr-hv: uploaded image name to have architecture in it's name 2021-07-12 10:17:15 +02:00
Pavel Raiskup
d10cb77f9b copr-hv: upload images only to appropriate hypervisors 2021-07-12 10:07:46 +02:00
Brendan Early
84d69178f3 packages-static: change deploy strategy to recreate 2021-07-11 20:39:46 -05:00
Brendan Early
4af473e648 packages-static: update solr files 2021-07-11 20:33:20 -05:00
Kevin Fenzi
6d8667f4c1 osbs_stg: also define fedora_required_images for stg
Signed-off-by: Kevin Fenzi <kevin@scrye.com>
2021-07-09 16:29:36 -07:00
Kevin Fenzi
b40ded4e40 osbs: comment out no longer available packages
Signed-off-by: Kevin Fenzi <kevin@scrye.com>
2021-07-09 16:06:57 -07:00
Kevin Fenzi
75b7a04207 osbs: playbook seems to want fedora_required_images defined.
Signed-off-by: Kevin Fenzi <kevin@scrye.com>
2021-07-09 15:45:07 -07:00
Kevin Fenzi
8b015fda21 inventory: re-add storinator to hardware
Signed-off-by: Kevin Fenzi <kevin@scrye.com>
2021-07-09 11:38:40 -07:00
Kevin Fenzi
c2fbb468e9 update appliedmicro hardware inventory
Signed-off-by: Kevin Fenzi <kevin@scrye.com>
2021-07-09 10:01:20 -07:00
Kevin Fenzi
908fb48c17 rdu-cc: Turns out the network there is a /24 not a /23
We only noticed this because we can't get to resources in the other /24
subnet from ours. :) All the other rdu-cc hosts are fine, just these
need adjusting.

Signed-off-by: Kevin Fenzi <kevin@scrye.com>
2021-07-09 08:52:53 -07:00
Kevin Fenzi
ef8456c46d storinator01: drop all the old communishift volumes.
Signed-off-by: Kevin Fenzi <kevin@scrye.com>
2021-07-09 07:10:53 -07:00
Kevin Fenzi
9b81171d17 storinator01: move the old config for exports to the new hostname
Signed-off-by: Kevin Fenzi <kevin@scrye.com>
2021-07-09 07:10:53 -07:00
Carl George
b74b272cbc Add mandatory days for EPEL Next 2021-07-09 13:38:24 +00:00
Kevin Fenzi
8a65238577 hardware: disable checking swap on hardware for now
RHEL8 kernels on virthosts in particular are aggressive about filling up
swap. This causes no issues, but is causing a lot of alert fatigue.
Ideally we would switch all swap checks to swap+memory free checks of
some sort. We should look at that when moving to zabbix.

Signed-off-by: Kevin Fenzi <kevin@scrye.com>
2021-07-09 05:56:55 -07:00
Kevin Fenzi
a42bb9e383 ipa/server: fix typo: yess to yes
Signed-off-by: Kevin Fenzi <kevin@scrye.com>
2021-07-08 09:45:41 -07:00
Kevin Fenzi
b39b6892fc vmhost-x86-08: add to inventory so it gets configured and updated instead of sitting idle and unrebooted
Signed-off-by: Kevin Fenzi <kevin@scrye.com>
2021-07-08 09:14:23 -07:00
Ryan Lerch
ff1395d99c ipsilon: clean up ipsilon role
This cleans up the ipsilon role, removing a bunch of old, commented out
lines, and removes files and templates that are no longer used.

Signed-off-by: Ryan Lerch <rlerch@redhat.com>
2021-07-08 10:13:41 +00:00
Kevin Fenzi
12e82cc7bc how many reverts can I do? This one should be reverted because we want it to not mess with resolv.conf
Revert "Revert "buildvm_s390x: try and not block ifcfg handling on s390x machines""

This reverts commit 290ef0841f.
2021-07-07 13:50:39 -07:00
Kevin Fenzi
290ef0841f Revert "buildvm_s390x: try and not block ifcfg handling on s390x machines"
This reverts commit 95eb10f492.

Turns out this is a rabbit hole we don't want to go down.
2021-07-07 13:49:25 -07:00
Kevin Fenzi
5d6f460b42 buildvm_s390x: the z/vm s390x instances use enc900 as network interface
Signed-off-by: Kevin Fenzi <kevin@scrye.com>
2021-07-07 13:43:22 -07:00
Kevin Fenzi
ad91eae586 base: adjust grouping to give the desired result in when for NM tweaking
Signed-off-by: Kevin Fenzi <kevin@scrye.com>
2021-07-07 12:56:46 -07:00
Kevin Fenzi
95eb10f492 buildvm_s390x: try and not block ifcfg handling on s390x machines
Signed-off-by: Kevin Fenzi <kevin@scrye.com>
2021-07-07 11:12:21 -07:00