[GH-ISSUE #40] Start update Manager on all Cluster bugs #33

New issue

Closed

opened 2026-03-02 15:47:22 +03:00 by kerem · 2 comments

kerem commented

2026-03-02 15:47:22 +03:00

Owner

Originally created by @ThibautBascoulergue on GitHub (Feb 6, 2026).
Original GitHub issue: https://github.com/PegaProx/project-pegaprox/issues/40

Hello,

Thank you again for the great work on PegaProx.
We have installed it and connected it to a cluster with 4 nodes + quorum + Ceph. Everything is correctly detected and working perfectly.

We tested the Update Manager module and noticed a few minor issues.

First issue: after launching Update Manager, it puts the first node into maintenance mode. The node is supposed to evacuate all VMs before completing maintenance. Some of our VMs failed to migrate (we currently have an open ticket with Proxmox), and we ended up with two or three VMs in a failed to migrate state.
Despite this, the node still proceeded with installing updates and rebooted.

I think it would be useful to have an option to pause the update process, allowing the user to manually migrate or shut down the remaining VMs, with full manual control, instead of the Update Manager forcing the process without any way to say “stop” 🙂

Second issue: once the first node finished installing its packages and initiated the reboot, the Update Manager immediately moved on to the next node without waiting for the first one to come back online (this may be expected behavior). However, it might be worth giving the user the choice here as well.

As a result, when the second node entered maintenance mode, it attempted to migrate VMs to the node that was still rebooting… which was quite surprising. Some VMs therefore had to be migrated to other online nodes to avoid a reboot with running VMs.

That’s all for now. I hope this feedback helps you move forward with this great project.
We will continue testing, and we will most likely add another, larger cluster soon.

Originally created by @ThibautBascoulergue on GitHub (Feb 6, 2026). Original GitHub issue: https://github.com/PegaProx/project-pegaprox/issues/40 Hello, Thank you again for the great work on PegaProx. We have installed it and connected it to a cluster with 4 nodes + quorum + Ceph. Everything is correctly detected and working perfectly. We tested the Update Manager module and noticed a few minor issues. First issue: after launching Update Manager, it puts the first node into maintenance mode. The node is supposed to evacuate all VMs before completing maintenance. Some of our VMs failed to migrate (we currently have an open ticket with Proxmox), and we ended up with two or three VMs in a failed to migrate state. Despite this, the node still proceeded with installing updates and rebooted. I think it would be useful to have an option to pause the update process, allowing the user to manually migrate or shut down the remaining VMs, with full manual control, instead of the Update Manager forcing the process without any way to say “stop” 🙂 Second issue: once the first node finished installing its packages and initiated the reboot, the Update Manager immediately moved on to the next node without waiting for the first one to come back online (this may be expected behavior). However, it might be worth giving the user the choice here as well. As a result, when the second node entered maintenance mode, it attempted to migrate VMs to the node that was still rebooting… which was quite surprising. Some VMs therefore had to be migrated to other online nodes to avoid a reboot with running VMs. That’s all for now. I hope this feedback helps you move forward with this great project. We will continue testing, and we will most likely add another, larger cluster soon.

kerem closed this issue

2026-03-02 15:47:22 +03:00

kerem commented

2026-03-02 15:47:23 +03:00

Author

Owner

@mkellermann97 commented on GitHub (Feb 8, 2026):

Hi @ThibautBascoulergue ,

We've implemented a fix for this in the latest version.

Regards,
Marcus

@mkellermann97 commented on GitHub (Feb 8, 2026): Hi @ThibautBascoulergue , We've implemented a fix for this in the latest version. Regards, Marcus

kerem commented

2026-03-02 15:47:23 +03:00

Author

Owner

@mkellermann97 commented on GitHub (Feb 8, 2026):

Hi @ThibautBascoulergue ,

please update PegaProx to the newest version.

Regards,
marcus

@mkellermann97 commented on GitHub (Feb 8, 2026): Hi @ThibautBascoulergue , please update PegaProx to the newest version. Regards, marcus