Jira (PUP-11588) dnfmodule issues with Nvidia CUDA repository

[Aria Li (Jira)]

Title: Message Title

Aria Li commented on PUP-11588   Re: dnfmodule issues with Nvidia CUDA repository We are unable to resolve this issue because the provider, dnfmodule, does not support the :upgradeable option because we cannot determine a latest version. Therefore, ensure => latest cannot be used for dnfmodule. For more information, please see this comment. We were able to reproduce this issue on a RedHat 8 VM with Puppet 7.18.0.  We began by checking if the latest stream of nvidia-driver could be manually installed: [root@faulty-believer ~]# sudo dnf module install nvidia-driver:latest ... Complete! [root@faulty-believer ~]# dnf module list --installed cuda-rhel8-x86_64 Name                         Stream                   Profiles                                 nvidia-driver                latest [e]               fm, src, ks, default [d] [i]             Hint: [d]efault, [e]nabled, [x]disabled, [i]nstalled [root@faulty-believer ~]# puppet resource package nvidia-driver package { 'nvidia-driver':   ensure   => '3:515.65.01-1.el8',   provider => 'dnf', } Then, we checked in stream 470 and 510 }}could be installed.{{  manifest.pp: package { 'nvidia-driver':   ensure   => '470',   provider => 'dnfmodule', }   [root@faulty-believer ~]# puppet apply manifest.pp Notice: /Stage[main]/Main/Package[nvidia-driver]/ensure: ensure changed 'latest' to '470' Notice: Applied catalog in 90.60 seconds manifest.pp: package { 'nvidia-driver':   ensure   => '510',   provider => 'dnfmodule', }  [root@faulty-believer ~]# puppet apply manifest.pp Notice: Compiled catalog for faulty-believer.delivery.puppetlabs.net in environment production in 0.31 seconds Notice: /Stage[main]/Main/Package[nvidia-driver]/ensure: ensure changed '470' to '510' Notice: Applied catalog in 107.34 seconds Then we used this manifest (manifest.pp) to attempt to install the latest stream: package { 'nvidia-driver':   ensure   => 'latest',   provider => 'dnf', } And got this: [root@faulty-believer ~]# puppet apply manifest.pp Notice: Compiled catalog for faulty-believer.delivery.puppetlabs.net in environment production in 0.32 seconds Notice: Applied catalog in 7.24 seconds [root@faulty-believer ~]# dnf module list --installed cuda-rhel8-x86_64 Name                          Stream                  Profiles                                     nvidia-driver                 510 [e]                 fm, src, ks, default [d] [i] We tried applying the same manifest with --debug to see what's happening: [root@faulty-believer ~]# puppet apply --debug manifest.pp ... Debug: Prefetching dnfmodule resources for package Debug: Executing: '/usr/bin/dnf --version' Debug: Executing: '/usr/bin/dnf module list -d 0 -e 1' Debug: Executing: '/usr/bin/dnf check-update' Debug: Package[nvidia-driver](provider=dnfmodule): Yum didn't find updates, current version (510) is the latest Debug: Finishing transaction 12040 Debug: Storing state Debug: Pruned old state cache entries in 0.00 seconds Since dnfmodule does not have its own defined latest, Puppet is calling the latest method from a superclass of dnfmodule and thinks there is no later version. Puppet thinks ensure => 'latest' is ensure => latest instead. More information can be found here.  Add Comment   This message was sent by Atlassian Jira (v8.20.11#820011-sha1:0629dd8) -- You received this message because you are subscribed to the Google Groups "Puppet Bugs" group. To unsubscribe from this group and stop receiving emails from it, send an email to puppet-bugs+unsubscr...@googlegroups.com. To view this discussion on the web visit https://groups.google.com/d/msgid/puppet-bugs/JIRA.454330.1657121517000.19731.1661213280058%40Atlassian.JIRA.