Bug 1894349

Summary:	yum update error messages
Product:	Red Hat Enterprise Linux 8	Reporter:	Martin <gmartins>
Component:	yum	Assignee:	Packaging Maintenance Team <packaging-team-maint>
Status:	CLOSED DUPLICATE	QA Contact:	swm-qe
Severity:	high	Docs Contact:
Priority:	unspecified
Version:	8.3	CC:	james.antill
Target Milestone:	rc
Target Release:	8.0
Hardware:	x86_64
OS:	Linux
Whiteboard:
Fixed In Version:		Doc Type:	If docs needed, set a value
Doc Text:		Story Points:	---
Clone Of:		Environment:
Last Closed:	2020-11-18 13:02:31 UTC	Type:	Bug
Regression:	---	Mount Type:	---
Documentation:	---	CRM:
Verified Versions:		Category:	---
oVirt Team:	---	RHEL 7.3 requirements from Atomic Host:
Cloudforms Team:	---	Target Upstream Version:
Embargoed:

Description Martin 2020-11-04 05:40:50 UTC

Description of problem:

Version-Release number of selected component (if applicable): RHEL 8.3
Linux [REDACTED] 4.18.0-193.28.1.el8_2.x86_64 #1 SMP Fri Oct 16 13:38:49 EDT 2020 x86_64 x86_64 x86_64 GNU/Linux

How reproducible:
run yum update

Steps to Reproduce:
1. yum update / upgrade

Actual results:

sudo yum upgrade
Updating Subscription Management repositories.
/usr/local/lib/python3.6/site-packages/dateutil/parser/_parser.py:70: UnicodeWarning: decode() called on unicode string, see https://bugzilla.redhat.com/show_bug.cgi?id=1693751
instream = instream.decode()

Last metadata expiration check: 0:04:24 ago on Tue 03 Nov 2020 11:25:45 PM CST.
Module yaml error: Artifact 'cuda-0:drivers-440.33.01-1.x86_64' was not in valid N-E:V-R.A format.
Module yaml error: Artifact 'cuda-0:drivers-440.33.01-1.x86_64' was not in valid N-E:V-R.A format.
Module yaml error: Artifact 'cuda-0:drivers-440.33.01-1.x86_64' was not in valid N-E:V-R.A format.
Module yaml error: Artifact 'cuda-0:drivers-440.33.01-1.x86_64' was not in valid N-E:V-R.A format.
Module yaml error: Artifact 'cuda-0:drivers-440.33.01-1.x86_64' was not in valid N-E:V-R.A format.
Module yaml error: Artifact 'cuda-0:drivers-440.33.01-1.x86_64' was not in valid N-E:V-R.A format.
Module yaml error: Artifact 'cuda-0:drivers-440.33.01-1.x86_64' was not in valid N-E:V-R.A format.
Dependencies resolved.
Nothing to do.
Complete!

Expected results:

sudo yum upgrade
Updating Subscription Management repositories.
Dependencies resolved.
Nothing to do.
Complete!

Additional info:

The error below is there for years. The system is already on py 3.8.5, but yum is still on 3.6, can't understand

/usr/local/lib/python3.6/site-packages/dateutil/parser/_parser.py:70: UnicodeWarning: decode() called on unicode string,

Comment 1 Radek Bíba 2020-11-05 07:05:27 UTC

Likely a duplicate of https://bugzilla.redhat.com/show_bug.cgi?id=1840142. Do you have the version of python3-dateutil mentioned in that bug?

Comment 2 Martin 2020-11-12 20:20:36 UTC

yes, thanks a lot, I manually fixed the python code there

python3-dateutil.noarch                             1:2.6.1-6.el8                                  @anaconda

what about:

"Module yaml error: Artifact 'cuda-0:drivers-440.33.01-1.x86_64' was not in valid N-E:V-R.A format."

Any idea? thanks,

M

Comment 3 Radek Bíba 2020-11-13 06:39:03 UTC

Not sure about the yaml error. What's the repo that triggers it? I tried https://developer.download.nvidia.com/compute/cuda/repos/rhel8/x86_64/ (just a guess based on the package name) and didn't run into that issue. Was the modular repodata file fixed in the meantime or is the error still reproducible?

FWIW:

# zgrep cuda /var/cache/dnf/nvidia-cde02fcb043171b7/repodata/37a3c329f829f14671b208b1d1546d492a165243e604626a4015d11978e1c56c-modules.yaml.gz | grep 440.33.01-1
            - nvidia-driver-cuda-3:440.33.01-1.el8.x86_64
            - nvidia-driver-cuda-libs-3:440.33.01-1.el8.x86_64
            - cuda-drivers-0:440.33.01-1.x86_64
            - nvidia-driver-cuda-3:440.33.01-1.el8.x86_64
            - nvidia-driver-cuda-libs-3:440.33.01-1.el8.x86_64
            - cuda-drivers-0:440.33.01-1.x86_64

ie. all the NEVRAs look good. It's true that cuda-0:drivers-440.33.01-1.x86_64 looks malformed, but I wonder where it comes from.

Comment 4 Martin 2020-11-13 22:04:32 UTC

This is my /var/cache/dnf
[root@XXX dnf]# zgrep cuda ./cuda-rhel8-x86_64-f1d7a46f058da57c/repodata/37a3c329f829f14671b208b1d1546d492a165243e604626a4015d11978e1c56c-modules.yaml.gz | grep 440.33.01-1.x86_64
            - cuda-drivers-0:440.33.01-1.x86_64
            - cuda-drivers-0:440.33.01-1.x86_64
[root@XXX dnf]# zgrep cuda ./cuda-10-2-local-10.2.89-440.33.01-739e0d407a724bcf/repodata/edfcc6a05a8445300758bc22904df983e731bc8a4fff4423b9e16c1055f9966e-modules.yaml.gz | grep 440.33.01-1
            - nvidia-driver-cuda-3:440.33.01-1.el8.x86_64
            - nvidia-driver-cuda-libs-3:440.33.01-1.el8.x86_64
            - cuda-0:drivers-440.33.01-1.x86_64
            - nvidia-driver-cuda-3:440.33.01-1.el8.x86_64
            - nvidia-driver-cuda-libs-3:440.33.01-1.el8.x86_64
            - cuda-0:drivers-440.33.01-1.x86_64
            - nvidia-driver-cuda-3:440.33.01-1.el8.x86_64
            - nvidia-driver-cuda-libs-3:440.33.01-1.el8.x86_64
            - cuda-0:drivers-440.33.01-1.x86_64

The version has been upgraded to 455.32.00-1.el8 and I don't think I need 440.33 anymore:

yum list installed | grep 440.33.01-1.x86_64
dnf list installed | grep 440.33.01-1.x86_64
(both empty)

How can I get rid of the message? thanks, M

Comment 5 Radek Bíba 2020-11-16 10:23:10 UTC

If the definition is malformed in the modularity file for the repo named "cuda-10-2-local-10.2.89-440.33.01", then dnf behaves as expected: it tells you about the issue. The error message can probably be silenced, but please contact the owners of the cuda-10-2-local-10.2.89-440.33.01 repo to have the repodata fixed there. Or if it's an abandoned repo and you have the packages from a different one now (cuda-rhel8-x86_64?), delete the repo file for the unneeded repo or disable the repo. You may need to remove the cache directories to make sure dnf doesn't see the broken definition anymore.

Comment 6 Martin 2020-11-16 21:43:54 UTC

cool, thanks, `rm /etc/yum.repos.d/cuda-10-2-local-10.2.89-440.33.01.repo` did the whole job. Safe to close ticket now.

Comment 7 Radek Bíba 2020-11-18 13:02:31 UTC

Glad it worked! I'm now closing this bug as a duplicate of 1840142, where the Unicode warning is tracked.

*** This bug has been marked as a duplicate of bug 1840142 ***