Ostreecontainer bootc test with reboot #1116

rvykydal · 2024-03-27T13:24:10Z

This adds a few modifications to the test added by @jikortus in #1096

rvykydal · 2024-03-27T13:25:08Z

I'll run the tests on the PR ('/test-platforms') when our runners are back healthy.

rvykydal · 2024-03-27T13:58:18Z

/test-platforms

rvykydal · 2024-04-02T10:54:59Z

/test-platforms

rvykydal · 2024-04-02T12:10:11Z

Apparently the test is failing on daily-iso (https://github.com/rhinstaller/kickstart-tests/actions/runs/8521608897/job/23340112174). The reboot of installed system fails.
From a local run:

kstest.log
virt-install.log

The iso is created for the test in https://github.com/rhinstaller/kickstart-tests/actions/workflows/daily-boot-iso-rawhide.yml. Perhaps there is some unexpected ID o label somewhere breaking some automation/defaults.

@jikortus any idea or hint?

jikortus · 2024-04-02T13:23:36Z

No idea right now, @rvykydal, but I'll execute the test with daily ISO and try to investigate what goes wrong.

jikortus · 2024-04-02T19:01:07Z

@rvykydal, I found out where the problem is. The fedora-bootc:eln image doesn't (naturally) contain any support for btrfs, which is used as Fedora's default scheme/FS. AFAIK there are only ELN and CentOS container images available (ATM?), so we need to workaround this by using a specific autopart scheme to avoid btrfs, at least until there's a proper Rawhide version of the container image which would include support for btrfs.

You can replace my previous commit containing the new test with jikortus@e45c825.

rvykydal · 2024-04-03T07:39:06Z

Great, thank you, indeed we are detecting different profiles for rawhide and daily-iso.

daily-iso:

12:05:22,126 DEBUG anaconda:anaconda: core.configuration.profile: Detecting a profile for ID=fedora, VARIANT_ID=None.
12:05:22,126 INFO anaconda:anaconda: core.configuration.profile: The 'fedora' profile is detected.
12:05:22,126 INFO anaconda:anaconda: core.configuration.anaconda: Load the 'fedora' profile configuration.

[Storage]
default_scheme = BTRFS

rawhide:

12:29:44,050 DEBUG anaconda:anaconda: core.configuration.profile: Detecting a profile for ID=fedora, VARIANT_ID=server.
12:29:44,050 INFO anaconda:anaconda: core.configuration.profile: The 'fedora-server' profile is detected.
12:29:44,050 INFO anaconda:anaconda: core.configuration.anaconda: Load the 'fedora-server' profile configuration.

[Storage]
file_system_type = xfs
default_scheme = LVM
default_partitioning =
    /    (min 2 GiB, max 15 GiB)

I think we should address it separately (#1120) and update this PR to make sure btrfs is not used on Fedora, similar as you suggest.

rvykydal · 2024-04-03T08:56:32Z

You can replace my previous commit containing the new test with jikortus@e45c825.

I've added a commit where I rather keep default autopartitioning scheme for rhel and enforce lvm only for Fedora. It is a bit too much of copying but we can improve it here: #1121

rvykydal · 2024-04-03T09:01:00Z

/test-platforms

jikortus · 2024-04-03T09:19:51Z

OK, that makes sense 👍. I hope the test will pass now at last.

jikortus · 2024-04-03T10:27:38Z

rpm-ostree-container-bootc.ks.in

@@ -3,7 +3,10 @@
 # depends on the referenced ostree container being bootable

 # Use the default settings.
-%ksappend common/common_no_payload.ks
+%ksappend common/common_no_storage_and_payload.ks
+%ksappend payload/default_packages.ks


Including a %packages section contradicts with a container image payload, at least for now (even though it may just be ignored by Anaconda).

Ah, good catch, thank you!

BTW, another thought - if I understand correctly the system of fragments, the platform-specific ones should apply if they exist and if they don't, a 'shared' one should apply, right? If this assumption is correct, shouldn't we define just one 'shared' and one specific, with LVM scheme for fedora_rawhide, leaving behind the other specific ones (rhelX) that would actually be covered by the 'shared' one?

ah, that would be great, let me check

You are right, thank you, updated.

rvykydal · 2024-04-03T10:33:38Z

/test-platforms

rvykydal · 2024-04-03T11:34:18Z

/test-platforms

rvykydal · 2024-04-03T13:25:32Z

/test-platforms

rvykydal · 2024-04-03T13:34:49Z

Download of boot.iso fails in test-platfroms for daily-iso. I'll give it a try tomorrow.

rvykydal · 2024-04-04T06:50:56Z

/test-platforms

rvykydal · 2024-04-04T07:04:48Z

rpm-ostree-container-bootc.sh

+}
+
+get_timeout() {
+   echo "80"


In my experience this test is pretty fast (not regenerating initramfs), less then 5 minutes. So instead of increasing the value from default 60 (mins) to 80, I'd rather keep it to 60 or even cut it down to 30? What do you think @jikortus ?

No problem at all, @rvykydal. The specific value was actually part of the initial work by Vladimir Slavik, so I thought there was a special reason for that. I retested it with get_timeout() removed and the test passed, so it is most likely not needed.

The ostree environments have most of the filesystem read-only (including /usr), but this can be overcome via an overlay. In such case the test script intended for run in the system resides in /var/lib/extensions/kickstart-tests/usr/libexec instead of the usual location (/usr/libexec).

It tests installation of a bootable container image (with bootc/bootupd) and boot of the installed system, as well as --remote and --stateroot parameters of ostreecontainer kickstart command. It's based on previous work from Vladimir Slavik.

The same place as is used for rpm payload

… test Patch created based on hints by jikortus.

The test is actually pretty fast (not generating initramfs) so there is no need to bigger timeout.

rvykydal · 2024-04-04T08:17:26Z

/test-platforms

jkonecny12

LGTM!

rvykydal marked this pull request as draft March 27, 2024 13:24

rvykydal mentioned this pull request Mar 27, 2024

New test - rpm-ostree-container-bootc + related improvements #1096

Closed

jikortus mentioned this pull request Mar 28, 2024

New test: rpm-ostree-container-luks #1118

Closed

rvykydal marked this pull request as ready for review April 2, 2024 10:58

rvykydal requested review from jstodola and jkonecny12 April 2, 2024 11:05

rvykydal changed the title ~~Ostreecontainer bootc update~~ Ostreecontainer bootc test with reboot Apr 2, 2024

jikortus mentioned this pull request Apr 2, 2024

New test: rpm-ostree-container-luks #1119

Merged

rvykydal mentioned this pull request Apr 3, 2024

daily-iso boot.iso uses "fedora" profile while rawhide boot.iso uses "fedora-server" #1120

Closed

rvykydal force-pushed the ostreecontainer-bootc-update branch from b409f3d to d6f7a47 Compare April 3, 2024 08:58

jikortus reviewed Apr 3, 2024

View reviewed changes

rvykydal force-pushed the ostreecontainer-bootc-update branch from d6f7a47 to 17c963a Compare April 3, 2024 10:32

rvykydal requested a review from jikortus April 3, 2024 11:11

rvykydal force-pushed the ostreecontainer-bootc-update branch 2 times, most recently from 3badcf7 to 2268a83 Compare April 3, 2024 11:32

rvykydal commented Apr 4, 2024

View reviewed changes

jikortus and others added 6 commits April 4, 2024 10:14

New test: rpm-ostree-container-bootc

c960efe

It tests installation of a bootable container image (with bootc/bootupd) and boot of the installed system, as well as --remote and --stateroot parameters of ostreecontainer kickstart command. It's based on previous work from Vladimir Slavik.

rpm-ostree-container-bootc: do the os substitutions only in .ks.in

aa25d11

Define per-platform ostreecontainer image in the single spot

4a9ba35

The same place as is used for rpm payload

Enforce non-btrfs autopart scheme on Fedora for ostreecontainer-bootc…

c927cb7

… test Patch created based on hints by jikortus.

Use default timeout for rpm-ostree-container-bootc

d7e66d1

The test is actually pretty fast (not generating initramfs) so there is no need to bigger timeout.

rvykydal force-pushed the ostreecontainer-bootc-update branch from 2268a83 to d7e66d1 Compare April 4, 2024 08:16

jkonecny12 approved these changes Apr 10, 2024

View reviewed changes

rvykydal merged commit b4d8aa0 into rhinstaller:master Apr 10, 2024
4 checks passed

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

Ostreecontainer bootc test with reboot #1116

Ostreecontainer bootc test with reboot #1116

rvykydal commented Mar 27, 2024 •

edited

Loading

rvykydal commented Mar 27, 2024

rvykydal commented Mar 27, 2024

rvykydal commented Apr 2, 2024

rvykydal commented Apr 2, 2024 •

edited

Loading

jikortus commented Apr 2, 2024

jikortus commented Apr 2, 2024

rvykydal commented Apr 3, 2024 •

edited

Loading

rvykydal commented Apr 3, 2024

rvykydal commented Apr 3, 2024

jikortus commented Apr 3, 2024

jikortus Apr 3, 2024 •

edited

Loading

rvykydal Apr 3, 2024

rvykydal Apr 3, 2024

jikortus Apr 3, 2024

rvykydal Apr 3, 2024 •

edited

Loading

rvykydal Apr 3, 2024

rvykydal commented Apr 3, 2024

rvykydal commented Apr 3, 2024

rvykydal commented Apr 3, 2024

rvykydal commented Apr 3, 2024

rvykydal commented Apr 4, 2024

rvykydal Apr 4, 2024 •

edited

Loading

jikortus Apr 4, 2024

rvykydal commented Apr 4, 2024

jkonecny12 left a comment

Ostreecontainer bootc test with reboot #1116

Ostreecontainer bootc test with reboot #1116

Conversation

rvykydal commented Mar 27, 2024 • edited Loading

rvykydal commented Mar 27, 2024

rvykydal commented Mar 27, 2024

rvykydal commented Apr 2, 2024

rvykydal commented Apr 2, 2024 • edited Loading

jikortus commented Apr 2, 2024

jikortus commented Apr 2, 2024

rvykydal commented Apr 3, 2024 • edited Loading

rvykydal commented Apr 3, 2024

rvykydal commented Apr 3, 2024

jikortus commented Apr 3, 2024

jikortus Apr 3, 2024 • edited Loading

Choose a reason for hiding this comment

rvykydal Apr 3, 2024

Choose a reason for hiding this comment

rvykydal Apr 3, 2024

Choose a reason for hiding this comment

jikortus Apr 3, 2024

Choose a reason for hiding this comment

rvykydal Apr 3, 2024 • edited Loading

Choose a reason for hiding this comment

rvykydal Apr 3, 2024

Choose a reason for hiding this comment

rvykydal commented Apr 3, 2024

rvykydal commented Apr 3, 2024

rvykydal commented Apr 3, 2024

rvykydal commented Apr 3, 2024

rvykydal commented Apr 4, 2024

rvykydal Apr 4, 2024 • edited Loading

Choose a reason for hiding this comment

jikortus Apr 4, 2024

Choose a reason for hiding this comment

rvykydal commented Apr 4, 2024

jkonecny12 left a comment

Choose a reason for hiding this comment

rvykydal commented Mar 27, 2024 •

edited

Loading

rvykydal commented Apr 2, 2024 •

edited

Loading

rvykydal commented Apr 3, 2024 •

edited

Loading

jikortus Apr 3, 2024 •

edited

Loading

rvykydal Apr 3, 2024 •

edited

Loading

rvykydal Apr 4, 2024 •

edited

Loading