Bug 1217872 - [SELinux] sedispatch fails to sent AVCs to system_bus_socket
Summary: [SELinux] sedispatch fails to sent AVCs to system_bus_socket
Status: RESOLVED WONTFIX
Alias: None
Product: PUBLIC SUSE Linux Enterprise Micro 5.5
Classification: openSUSE
Component: Base (show other bugs)
Version: unspecified
Hardware: Other Other
: P3 - Medium : Normal
Target Milestone: ---
Assignee: Zdenek Kubala
QA Contact: Jose Lausuch
URL:
Whiteboard:
Keywords:
Depends on:
Blocks:
 
Reported: 2023-12-07 10:32 UTC by Johannes Segitz
Modified: 2024-02-12 15:41 UTC (History)
5 users (show)

See Also:
Found By: ---
Services Priority:
Business Priority:
Blocker: ---
Marketing QA Status: ---
IT Deployment: ---
gyribeiro: needinfo? (jalausuch)


Attachments
Full journal (2.31 MB, text/plain)
2024-01-29 10:27 UTC, Jose Lausuch
Details
journal_alp (1.37 MB, text/plain)
2024-01-29 11:03 UTC, Jose Lausuch
Details

Note You need to log in before you can comment on or make changes to this bug.
Description Johannes Segitz 2023-12-07 10:32:43 UTC
https://openqa.suse.de/tests/12990879#

there's
Dec 07 10:06:59.153041 localhost.localdomain sedispatch[646]: Connection Error (Failed to connect to socket /run/dbus/system_bus_socket: Connection refused): AVC Will be dropped

I just stumbled across this as this triggers for SELinux denials because of the AVC. I'll soft-fail this for now. Probably something for Zdenek
Comment 1 Zdenek Kubala 2023-12-11 12:43:20 UTC
Hmm, it seems like a one time glitch. Following QA runs passed this stage.
https://openqa.suse.de/tests/12990879#next_previous.
Comment 3 Zdenek Kubala 2024-01-08 10:35:27 UTC
(In reply to Johannes Segitz from comment #0)
> https://openqa.suse.de/tests/12990879#
> 
> there's
> Dec 07 10:06:59.153041 localhost.localdomain sedispatch[646]: Connection
> Error (Failed to connect to socket /run/dbus/system_bus_socket: Connection
> refused): AVC Will be dropped
> 
> I just stumbled across this as this triggers for SELinux denials because of
> the AVC. I'll soft-fail this for now. Probably something for Zdenek

Unfortunately test has been deleted. Can you please provide link to actual test where it fails or failed in the past? Thank you
Comment 4 Johannes Segitz 2024-01-09 12:28:40 UTC
e.g. https://openqa.suse.de/tests/13210311#

seems to happen on every run, just check the journal. I don't know where it is triggered
Comment 5 Zdenek Kubala 2024-01-18 09:42:19 UTC
@johannes I have checked 5.5 tests in https://openqa.suse.de/group_overview/420?limit_builds=20&limit_builds=400 and I found probably initial error here  https://openqa.suse.de/tests/overview?distri=sle-micro&version=5.5&build=20231114-1&groupid=420 and from that time there is time to time occurrence in multiple tests of `DVD-update` flavour failing on `Connection Error (Failed to connect to socket /run/dbus/system_bus_socket: Connection refused): AVC Will be dropped` in arm or other architectures. But errors shows only in tests +- 1 month ago and from that time soft-errors are on 11 showing no sign of the error mentioned above. There are no logs available to debug in the OpenQA. 

It looks like a temp. issue probably resolved in updating another packages. Anyway I'm trying to use aarch64 asset image from the failed test to see if the issue will occure. Anyway we can keep the bug open in case it will reoccur or till I try to reproduce it in my aarch64 image.
Comment 6 Zdenek Kubala 2024-01-22 12:43:10 UTC
I wasn't able to reproduce the issue with the aarch64 image in the tests.
Comment 7 Johannes Segitz 2024-01-22 14:09:07 UTC
(In reply to Zdenek Kubala from comment #6)
Then it seems to be an issue that happens in openQA only. We had some of those before. Can you please debug via test cases in openQA and try to figure out the cause?
Comment 8 Zdenek Kubala 2024-01-25 09:42:32 UTC
I asked QA guys to get notification in this bug when that error occurs again so I can see full logs. They added an automatic trigger for that error. Meanwhile I'm looking in the test cases.
Comment 9 Jose Lausuch 2024-01-29 10:27:43 UTC
Created attachment 872251 [details]
Full journal

This is happening in every run for SLE Micro 5.5
e.g. 
https://openqa.suse.de/tests/13369903#step/journal_check/21

we have white listed the bug, so that's why you see it as "orange". I have attached the full journal of the test execution.
Comment 10 Jose Lausuch 2024-01-29 11:03:05 UTC
Created attachment 872252 [details]
journal_alp

This is also applicable to ALP:
https://openqa.suse.de/tests/13373368#step/journal_check/9
(see also attached journal)
Comment 11 Zdenek Kubala 2024-01-30 12:07:19 UTC
(In reply to Jose Lausuch from comment #9)
> Created attachment 872251 [details]
> Full journal
> 
> This is happening in every run for SLE Micro 5.5
> e.g. 
> https://openqa.suse.de/tests/13369903#step/journal_check/21
> 
> we have white listed the bug, so that's why you see it as "orange". I have
> attached the full journal of the test execution.

Hi Jose, your link leads to the error `plugin /usr/sbin/sedispatch terminated unexpectedly` and that is handled by bnc#1215377 as mentioned in "orange" slide. Or did you want to point out to something else?
Comment 13 Jose Lausuch 2024-02-12 15:41:58 UTC
You are right. I can't find this error message any longer. The only one I see related to sedispatch is https://bugzilla.suse.com/show_bug.cgi?id=1215377

Let me mark it as won't fix and re-open if we see it again.

Thank you.