[GH-ISSUE #22] ⁉️ Question: Rolling update without reboot evacuates/migrates VMs — is this intended? #18

New issue

Closed

opened 2026-03-02 15:47:14 +03:00 by kerem · 2 comments

kerem commented

2026-03-02 15:47:14 +03:00

Owner

Originally created by @PhiGi87 on GitHub (Jan 29, 2026).
Original GitHub issue: https://github.com/PegaProx/project-pegaprox/issues/22

Hey,

I have a question about the current rolling update behavior.

When I run a rolling update without reboot, the host still gets evacuated and the VMs are migrated away. Is that intended behavior?

I’m wondering if it makes sense in this case: updates that don’t require a restart are often safe for running VMs, so evacuating the host might be unnecessary overhead (extra migrations, time, load).

Would it make sense to make “evacuate host / migrate VMs” optional for non-reboot updates? This is just an idea/question — I’m not fully sure myself and wanted to understand the reasoning behind the current behavior.

Thanks!

Originally created by @PhiGi87 on GitHub (Jan 29, 2026). Original GitHub issue: https://github.com/PegaProx/project-pegaprox/issues/22 Hey, I have a question about the current rolling update behavior. When I run a rolling update *without* reboot, the host still gets evacuated and the VMs are migrated away. Is that intended behavior? I’m wondering if it makes sense in this case: updates that don’t require a restart are often safe for running VMs, so evacuating the host might be unnecessary overhead (extra migrations, time, load). Would it make sense to make “evacuate host / migrate VMs” optional for non-reboot updates? This is just an idea/question — I’m not fully sure myself and wanted to understand the reasoning behind the current behavior. Thanks!

kerem

2026-03-02 15:47:14 +03:00

closed this issue
added the
enhancement
label

kerem commented

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

Author

Owner

@mkellermann97 commented on GitHub (Jan 29, 2026):

Hi @PhiGi87 ,
Yes, this behavior is intended.
The idea behind it is to protect your VMs in case something goes wrong during the update – they'll keep running regardless. We've actually had cases where VMs crashed after a normal update without any major changes, so we want to be cautious here.
That said, we'll add an option to override this, though it will be marked as not recommended.

Regards,
Marcus

@mkellermann97 commented on GitHub (Jan 29, 2026): Hi @PhiGi87 , Yes, this behavior is intended. The idea behind it is to protect your VMs in case something goes wrong during the update – they'll keep running regardless. We've actually had cases where VMs crashed after a normal update without any major changes, so we want to be cautious here. That said, we'll add an option to override this, though it will be marked as not recommended. Regards, Marcus

kerem commented

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

Author

Owner

@MrMasterbay commented on GitHub (Jan 30, 2026):

Hi @PhiGi87,

a button is now introduced there.
I will close this issue now.

BEst regards,
Nico

@MrMasterbay commented on GitHub (Jan 30, 2026): Hi @PhiGi87, a button is now introduced there. I will close this issue now. BEst regards, Nico