devel December 2022

devel@lists.centos.org

28 participants
24 discussions

Discussion: Open technical questions concerning SIG content for RHEL releases
by Peter Georg 09 Jun '23

09 Jun '23

Dear all, There has been an inquiry to the CentOS Board to clarify how long SIGs can expect to be able to build against RHEL build targets [1]. This is necessary to allow SIGs themselves to communicate expected lifecycle of their content to users. The outcome of this, in summary, is that SIGs should be able to build against RHEL until it reaches EOL (end of Maintenance Phase) + 30 days (no access to ELS content). This policy follows the one used by EPEL. Note: It is (obviously) still up to every single SIG to decide if and for how long they want to support a particular RHEL (or CentOS Stream) release. This inquiry also included respectively raised some technical questions. These technical questions are not to be decided by the CentOS Board. In [2] it has been recommended to discuss these on centos-devel. So let's do that. I want to start with two open questions which I do know of. In case there are other open question concerning this matter please raise them here. First open question is where to push content produced by SIGs for RHEL8 to. To better understand the issue, let me summarize the current situation, for which we first need to define the used format for tags in CBS: <sig><el>-<project>-<version>-{candidate,testing,release} where <el> can currently be 7,8,8s,9,9s Packages tagged for -testing are pushed to: https://buildlogs.centos.org/centos/<el>/<sig>/<arch>/<project>-<version>/ Packages tagged for -released are push to: If el = 8 or 8s: debuginfo rpm packages: https://debuginfo.centos.org/centos/<el>/<sig>/<arch>/<project>-<version>/ src.rpm packages: https://vault.centos.org/centos/<el>/<sig>/Source/<project>-<version>/ Everything else: http://mirror.centos.org/centos/<el>/<sig>/<arch>/<project>-<version>/ If el = 9 or 9s: http://mirror.stream.centos.org/SIGs/<el>/<sig>/<arch>/<project>-<version>/ For src.rpm <arch> is source. Note: Although it is not relevant for the discussion, in fact on the mirrors the suffix "s" of <el> is replaced by "-stream". This currently is all fine. However it is planned that the infrastructure currently used for packages built for 8 or 8s will disappear once CentOS Stream 8 and CentOS Linux 7 are EOL (June 30th, 2024). So in case any SIG decides to produce content for RHEL 8 after that date packages can not be distributed properly anymore. Only using the -testing tag and thus https://buildlogs.centos.org is not really an option as it would probably produce too much load on these. My proposal: After CentOS Stream 8 went EOL (May 31st, 2024), which is the same date as the end of the Full support phase of RHEL 8, at least the -release tag of all RHEL 8 build targets is locked and SIGs have to opt-in in case they want to build content for RHEL 8 during its Maintenance Support phase (End: May 31st 2029) for a particular build target. For these, tagging for -release will then push content to http://mirror.stream.centos.org/SIGs/8/<sig>/<arch>/<project>-<version>/, i.e. the same location as used for 9 and 9s. This gives SIGs one month for all required changes before the currently infrastructure disappears. Of course this means additional work for SIGs and Infra. However at some point the location where to push content to has to be changed before the currently used infrastructure disappears. Choosing an earlier date would result in work for SIGs who do not want to continue providing any packages once RHEL 8 enters Maintenance Support phase anyway. The second open question concerns the centos-release-* packages provided by SIGs to allow users to easily consume SIGs' content. For 8s and 9s the CBS tags extras<el>-extras-common-{candidate,testing,release} are used to build these packages. This repository is added in CentOS Stream 8 and 9. For packages build for RHEL 8 and 9 there is currently no common way to provide any means of easing the process to consume SIGs' content. My proposal to fix this is by adding extras<el>-extras-common-{candidate,testing,release} for <el> = 8 and 9, i.e., using the same system as currently used for 8s and 9s. To further ease the process I propose to introduce a package named centos-release-extras which contains the repository config pointing to the content of the tags extras<el>-extras-common-{testing,release} (only -release enabled by default) and the CentOS-SIG-Extras GPG key. The centos-release-extras packages itself would be built in the extras<el>-extras-common-el<el> build target. Users of RHEL would then only need to install this single package to allow them to easily install any other centos-release-* packages. Obviously someone needs to maintain the centos-release-extras package. I volunteer to maintain this package. [1]: https://git.centos.org/centos/board/issue/82 [2]: https://pagure.io/centos-infra/issue/1002

4 11

Few questions on cloud images
by Pino Toscano 06 Jun '23

06 Jun '23

Hello, I have few questions related to the cloud images provides for c8s and c9s, i.e.: - https://cloud.centos.org/centos/8-stream/x86_64/images/ - https://cloud.centos.org/centos/9-stream/x86_64/images/ 1) I see the c9s images are regularly updated, while the c8s images not; is this wanted, or simply a mistake/miss somewhere? If so, will they be kept up-to-date just like the c9s images? 2) are the Vagrant images there officially supported? Let's say you use one, and find a bug; then check on a "classic" installation (baremetal/VM), and the bug isn't there; can users report problems in those cases? 3) would it be possible to have either unversioned or "-latest" symlinks for the latest versions of all the files (i.e. images, checksums, etc)? This way, it will be much easier for users to fetch the latest versions with no need to change URLs frequently. Of course, the assumption is that users of this are always willing to use the latest version. Thanks! -- Pino Toscano

7 14

Fedora/CentOS planned outage 2022-12-08 19:00 UTC
by Stephen Smoogen 10 Jan '23

10 Jan '23

lanned Outage - IAD2 Outage - 2022-12-08 19:00 UTC There will be an outage starting at 2022-12-08 19:00 UTC, which will last approximately 4 hours. To convert UTC to your local time, take a look at http://fedoraproject.org/wiki/Infrastructure/UTCHowto or run: date -d '2022-12-08 19:00 UTC' Reason for outage: There will be a multi hour outage as a hardware firewall is changed out of the IAD2 data center where most Fedora systems are housed. Outages should be in short cycles as the firewall is changed over to new hardware and rules are tested and confirmed. Affected Services: Most Fedora and CentOS Services will be affected Build systems for s390x will need to be restarted as NFS will break Builds in CentOS Stream and other infrastructure will be blocked during this time. Ticket Link: https://pagure.io/fedora-infrastructure/issue/NNNN Please join #fedora-admin or #fedora-noc on irc.libera.chat or add comments to the ticket for this outage above. -- Stephen Smoogen, Red Hat Automotive Let us be kind to one another, for most of us are fighting a hard battle. -- Ian MacClaren

2 5

Proposal: CentOS Public Board Meetings
by Rich Bowen 07 Jan '23

07 Jan '23

In January of 2021, the CentOS Board of Directors opened our monthly meetings to all SIG chairs, in a first step towards more transparent governance. We wanted to see how the larger audience changed the meeting dynamics, if at all, as well as welcoming a broader pool of voices. We are now ready for the second part of this process. Starting in the August meeting, we will be inviting the entire CentOS community to attend these meetings. Due to the nature of the Internet, we won’t be simply publishing the meeting URL, as this often leads to disruptive behavior in meetings, as I am sure many of you have experienced. Rather, in the week before the meeting, we’ll send, to this list, a way to request an invitation to the meeting. In this way, we hope to limit the incidence of such disruptions, while also making the meeting available to anyone who wishes to attend. We will be working on the rules of engagement for these meetings, but they will be, at minimum: * Wait to be recognized by the Chair before speaking * Respect the Chair when told your time to speak is over - this will allow us to remain on agenda, and complete the meetings in the allotted time * In the event that there are Board-confidential topics, these will be put at the end of the meeting, in Executive Session, and guests will be asked to leave. We hope to minimize these items, but they do sometimes happen. The most common scenarios in which this may happen are personnel issues, or information that Red Hat wishes to share with the Board, but is not yet public. * Muting of participants, or, in extreme situations, ejection from the meeting, is at the sole discretion of the Chair. * Meetings will be recorded, and published to YouTube (possibly with edits/redactions, as approved by the Directors). Thus, by joining the meeting you consent to have your presence at the meeting, and anything you say during the meeting, made public. Because there will seldom be time for Q&A during the meeting itself, we plan to add a post-meeting IRC office hours session for followup questions. This will probably be scheduled for a week after the board meeting, to allow time to publish meeting minutes We welcome your comments about the above proposal, and thank you all for your patience as we implement these changes to our governance process.

8 27

Alt Images SIG meeting minutes (2022/12/22)
by Troy Dawson 22 Dec '22

22 Dec '22

Meeting summary --------------- * Image build systems: livecd and kiwi - Infra ticket to create build targets in CBS https://pagure.io/centos-infra/issue/988 -- Looks like nothing is going to happen until next year. But it looks like it will be done before the ImageBuilder stuff happens. - koji is not importing all the comps stuff needed -- https://pagure.io/koji/issue/3626 -- koji skips environment and category groups when importing the comps files. -- The workaround it to use just regular groups for everything. It's alot more work, but at least it is possible. - Neal's backport of the upstreamed kiwi-live feature to support automatic persistent overlays on live USBs was merged into RHEL 9's dracut and it should be in c9s now: https://kojihub.stream.centos.org/koji/buildinfo?buildID=27844 -- Some anaconda patches/fixes are needed so this doing break installations from livecd's. * Image build systems: ImageBuilder - Infra ticket to create build targets in CBS https://pagure.io/centos-infra/issue/971 -- Not expected to get any movement from infra until next year. * Open Floor - Clarified that meetings are twice a month, on the 2nd and 4th Thursday -- Next meeting will be on January 12, 2023 at the usual time. People Present (lines said) --------------------------- * tdawson (66) * Eighth_Doctor (42) * davide (9) * davdunc[m (8) * salimma (6) * centguard (5) * rcallicotte (5) * michel-slm (0 Minutes and Log --------------- https://www.centos.org/minutes/2022/December/centos-meeting.2022-12-22-19.0… https://www.centos.org/minutes/2022/December/centos-meeting.2022-12-22-19.0… https://www.centos.org/minutes/2022/December/centos-meeting.2022-12-22-19.0…

1 0

Changes to CentOS CI: New OpenShift instance and Updates
by Camila Granella 22 Dec '22

22 Dec '22

*TL;DR: new OpenShift instance is up and running for CentOS CI, migrate your workloads by March 31st, 2023. Test your access to both the Duffy API and the new OpenShift instance by the end of 2022. CentOS CI changes approaching the end of Phase 3. * Hello everyone, We are happy to announce that the new OpenShift instance is up and running (ticket [0]). For the tenants that opted in to continue using CentOS CI, you are being required to *migrate your workload to the new OpenShift instance by the end of Q1CY23*, with the final deadline being *March 31st, 2023*. In order to support you through the migration process, the team has created a quick guide that can be found in [1]. We would also like to request you test your access to both the Duffy API and the new OpenShift instance by the *end of 2022*. Please reach out to us in case it is not working. Here is the link to the new OCP cluster [2]. As a reminder, this new OCP cluster is running in AWS, on top of EC2 instances, so coming with different limitations when compared with “on-premises bare-metal OCP deployments”. One notable change is that it is not possible to expose /dev/kvm and so there is no way for tenants to deploy VMs through OpenShift (like with the kubevirt operator or else). Tenants having a need to use VMs can then request some through Duffy API. Finally, here's a friendly reminder of the CentOS CI changes' current status. According to our previous communications [3], we are approaching the end of the last phase of the changes that have been implemented to CentOS CI. *Phase 3 - Decommission* - Legacy/compatibility API deprecated and requests (even for EC2 instances) will no longer be accepted - All tenants that opted in will be using only EC2 for aarch64/x86_64 and on-premise cloud for ppc64le For a full description of the project phases, you can check [3]. Please, feel free to reach out to us in case you have any questions or concerns. Regards, [0] https://pagure.io/centos-infra/issue/969 [1] https://sigs.centos.org/guide/ci/#migration-to-new-ci-instance [2] https://console-openshift-console.apps.ocp.cloud.ci.centos.org/ [3] https://lists.centos.org/pipermail/ci-users/2022-June/004547.html -- Camila Granella Associate Manager, Software Engineering Red Hat <https://www.redhat.com/> @Red Hat <https://twitter.com/redhat> Red Hat <https://www.linkedin.com/company/red-hat> Red Hat <https://www.facebook.com/RedHatInc> <https://www.redhat.com/>

2 1

CentOS Connect Schedule
by Shaun McCance 21 Dec '22

21 Dec '22

The full schedule for CentOS Connect at FOSDEM is up with 14 amazing talks. CentOS Connect is a free mini-conference on topics across the Enterprise Linux ecosystem. If you're coming to FOSDEM, come in a day early and join us. Registration is free but strongly recommended. Hope to see you there. https://connect.centos.org/

2 1

FreeIPA / Samba / SSSD - new packages same problems
by lejeczek 21 Dec '22

21 Dec '22

Hi gents. It seems this has been a routine - each time when either sssd or Samba or FreeIPA gets new releases (with freeIPA deployed) there will be 'rpm' conflicts: Problem 1: package sssd-common-pac-2.7.3-5.el8.x86_64 requires libndr.so.2()(64bit), but none of the providers can be installed - package sssd-common-pac-2.7.3-5.el8.x86_64 requires libndr.so.2(NDR_0.0.1)(64bit), but none of the providers can be installed - cannot install both samba-client-libs-4.17.2-2.el8.x86_64 and samba-client-libs-4.16.4-2.el8.x86_64 - cannot install both samba-client-libs-4.17.2-2.el8.x86_64 and samba-client-libs-4.15.3-0.el8.x86_64 - cannot install both samba-client-libs-4.17.2-2.el8.x86_64 and samba-client-libs-4.15.4-0.el8.x86_64 - cannot install both samba-client-libs-4.17.2-2.el8.x86_64 and samba-client-libs-4.15.5-0.el8.x86_64 - cannot install both samba-client-libs-4.17.2-2.el8.x86_64 and samba-client-libs-4.15.5-3.el8.x86_64 - cannot install both samba-client-libs-4.17.2-2.el8.x86_64 and samba-client-libs-4.15.5-4.el8.x86_64 - cannot install both samba-client-libs-4.17.2-2.el8.x86_64 and samba-client-libs-4.15.5-5.el8.x86_64 - cannot install both samba-client-libs-4.17.2-2.el8.x86_64 and samba-client-libs-4.15.5-8.el8.x86_64 - cannot install both samba-client-libs-4.17.2-2.el8.x86_64 and samba-client-libs-4.16.1-0.el8.x86_64 - cannot install both samba-client-libs-4.17.2-2.el8.x86_64 and samba-client-libs-4.16.2-1.el8.x86_64 - cannot install both samba-client-libs-4.17.2-2.el8.x86_64 and samba-client-libs-4.16.4-1.el8.x86_64 - cannot install the best update candidate for package sssd-common-pac-2.7.3-4.el8.x86_64 - cannot install the best update candidate for package samba-client-libs-4.16.4-2.el8.x86_64 Problem 2: package sssd-idp-2.7.3-4.el8.x86_64 requires sssd-common = 2.7.3-4.el8, but none of the providers can be installed - package libsss_autofs-2.7.3-5.el8.x86_64 conflicts with sssd-common < 2.7.3-5.el8 provided by sssd-common-2.7.3-4.el8.x86_64 - cannot install the best update candidate for package sssd-idp-2.7.3-4.el8.x86_64 - cannot install the best update candidate for package libsss_autofs-2.7.3-4.el8.x86_64 Problem 3: package sssd-common-pac-2.7.3-5.el8.x86_64 requires libndr.so.2()(64bit), but none of the providers can be installed - package sssd-common-pac-2.7.3-5.el8.x86_64 requires libndr.so.2(NDR_0.0.1)(64bit), but none of the providers can be installed Or it's just me? many thanks, L.

6 10

dnf stuff has broken dependencies
by Leon Fauster 19 Dec '22

19 Dec '22

FYI: The samba issue is gone but following is still an issue: # LANG=C yum update Last metadata expiration check: 0:00:18 ago on Fri Nov 25 23:48:00 2022. Dependencies resolved. Problem 1: package python3-dnf-plugin-modulesync-4.0.21-14.1.el8.noarch requires python3-dnf-plugins-core = 4.0.21-14.1.el8, but none of the providers can be installed - cannot install both python3-dnf-plugins-core-4.0.21-16.el8.noarch and python3-dnf-plugins-core-4.0.21-14.1.el8.noarch - cannot install the best update candidate for package python3-dnf-plugins-core-4.0.21-14.1.el8.noarch - cannot install the best update candidate for package python3-dnf-plugin-modulesync-4.0.21-14.1.el8.noarch Problem 2: problem with installed package python3-dnf-plugin-modulesync-4.0.21-14.1.el8.noarch - package python3-dnf-plugin-modulesync-4.0.21-14.1.el8.noarch requires python3-dnf-plugins-core = 4.0.21-14.1.el8, but none of the providers can be installed - cannot install both python3-dnf-plugins-core-4.0.21-16.el8.noarch and python3-dnf-plugins-core-4.0.21-14.1.el8.noarch - package dnf-plugins-core-4.0.21-16.el8.noarch requires python3-dnf-plugins-core = 4.0.21-16.el8, but none of the providers can be installed - cannot install the best update candidate for package dnf-plugins-core-4.0.21-14.1.el8.noarch ============================================================================================================================================================================= Package Architecture Version Repository Size ============================================================================================================================================================================= Skipping packages with conflicts: (add '--best --allowerasing' to command line to force their upgrade): python3-dnf-plugins-core noarch 4.0.21-16.el8 baseos 258 k Skipping packages with broken dependencies: dnf-plugins-core noarch 4.0.21-16.el8 baseos 75 k Transaction Summary ============================================================================================================================================================================= Skip 2 Packages Nothing to do. Complete! -- Leon

2 2

CPE Weekly Update – Week 50 2022
by Lenka Segura 16 Dec '22

16 Dec '22

*Hi everyone,* *Reminder: This is the last email we are sending! If you want to stay in touch with the CPE updates, please follow the instructions at the bottom.* *This is a weekly report from the CPE (Community Platform Engineering) Team.The report could be found at https://communityblog.fedoraproject.org/cpe-weekly-update-week-50-2022/ <https://communityblog.fedoraproject.org/cpe-weekly-update-week-50-2022/>If you want to receive weekly reports by email in the future, please subscribe to either https://communityblog.fedoraproject.org/ <https://communityblog.fedoraproject.org/> or https://discussion.fedoraproject.org/c/news/commblog/61 <https://discussion.fedoraproject.org/c/news/commblog/61>. We will stop sending them in 2023.Regards,CPE Team*

2 1

2024

2023

2022

2021

2020

2019

2018

2017

2016

2015

2014

2013

2012

2011

2010

2009

2008

2007

2006

2005

devel December 2022