Node maintenance mode UI

chrispage1

Member
Sep 1, 2021
86
39
23
32
Hi, the new node maintenance mode is a great addition to Proxmox, thank you!

Running sudo ha-manager crm-command node-maintenance enable pve01 worked perfectly.

However, it'd be really nice to be able to manage maintenance state directly from Datacenter > HA. Is this something that's in the works?

It'd also be really nice to see nodes under maintenance mode with a maintenance flag on the sidebar. Something like the below :)

1681375169984.png
 
Maybe off-topic: why sudo? Normally, there are no other os user accounts on the machine.

Only because I'm logged into the box with my own username that has sudoers permissions :) I prefer not to login directly as root... Just in case!
 
Yep that's fair enough. It breaks autocomplete if it doesn't have access to what you are calling. For me its just personal preference, makes me think twice about the command I'm entering. For example, if I accidentally typed rm -rf / as my user, worse things have happened. Whereas if I accidentally did that as root (say, I meant to delete the current folder rm -rf ./) I'd quickly be up s**t creek without a paddle :)

So by typing sudo it makes me extremely aware of what I'm typing. Granted - there are cases where it's just easier to be escalated to root and I often do that too.
 
This is an important thread that I’ll add to.

LnxBil is making a valid point. It’s not compliant from an enterprise perspective to run as root so routinely like we do in proxmox. Don’t get me wrong I love the convenience of clicking once and being logged in as root like that but I get why he’s working that way. I don’t think that’s a big issue overall but that is something to think about. A hacker breaking Into our proxmox sessions is one click away from root access.

Last night I too ran ha-manager crm-command node-maintenance enable proxmox3 on my third node and I was *blown away!* Literally giddy with laughter. My cluster is only 3 weeks old so it’s all still new and exciting but that maintenance mode is *fantastic!*. I was watching a film through a browser on my Windows VM. The only reason I knew a node went down and migrated the VM to another node is the film which I was watching on MS RD on an iPad via Wi-Fi burped for less than 1 second and then continued playback without me having to do anything. The ZFS pool and cluster worked flawlessly! Im not even using EC ram! This shouldn’t even be working. I’m blown away.

The only thing missing is we need this in the GUI and we need to be able to schedule maintenance without having to write our own scripts by hand.

Proxmox team if you’re reading this, fantastic work on maintenance mode. I really hope it gets more improvements.
 
Hi, the new node maintenance mode is a great addition to Proxmox, thank you!

Running sudo ha-manager crm-command node-maintenance enable pve01 worked perfectly.

However, it'd be really nice to be able to manage maintenance state directly from Datacenter > HA. Is this something that's in the works?

It'd also be really nice to see nodes under maintenance mode with a maintenance flag on the sidebar. Something like the below :)

View attachment 49161
Agree this needs to be added to the GUI ASAP so Proxmox can become more mainstream

It shouldn't be that hard.
 
And (temporarily) enabling maintenance mode for a server should be integrated in the shutdown procedures too.

Resources defined as HA resources shouldn't be stopped when a server goes down but migrated to another available server the same kind as when setting the server manually in maintenance mode before shutting down it.
 
Last edited:
And (temporarily) enabling maintenance mode for a server should be integrated in the shutdown procedures too.

Resources defined as HA resources shouldn't be stopped when a server goes down but migrated to another available server the same kind as when setting the server manually in maintenance mode before shutting down it.

I always enable maintenance mode just in case but I thought failover already automatically occurred for HA assets that had additional nodes defined with lower priority.
 
I always enable maintenance mode just in case but I thought failover already automatically occurred for HA assets that had additional nodes defined with lower priority.
Yes, according to the documentation a failover will happen at some point later, but the resources are shutdown first, that means there is an interruption in service.
I've found a point in the documentation later stating that it's possible to set the shutdown policy to "migrate" (https://pve.proxmox.com/pve-docs/chapter-ha-manager.html#ha_manager_shutdown_policy) but there is no hint where to set it nor did I find such a functionality in the GUI :(
 
on the GUI:

Datacenter -> Options -> HA Settings
 
i++
=

I really need this!

btw.
I additionally use this command to stop all replication jobs:

On all Hosts in the Cluster
Disable/stop ALL
Code:
repjobs=$(pvesr list |  cut -f1 -d" " | tail -n +2);for i in $repjobs; do pvesr disable $i; done

Enable/start schedule with:
Code:
repjobs=$(pvesr list |  cut -f1 -d" " | tail -n +2);for i in $repjobs; do pvesr enable $i; done


On the current host
Disable/stop jobs on the current host
Code:
repjobs=$(pvesr status |  cut -f1 -d" " | tail -n +2);for i in $repjobs; do pvesr disable $i; done

Enable/start schedule jobs on the current host with:
Code:
repjobs=$(pvesr status |  cut -f1 -d" " | tail -n +2);for i in $repjobs; do pvesr enable $i; done
 
Last edited:

About

The Proxmox community has been around for many years and offers help and support for Proxmox VE, Proxmox Backup Server, and Proxmox Mail Gateway.
We think our community is one of the best thanks to people like you!

Get your subscription!

The Proxmox team works very hard to make sure you are running the best software and getting stable updates and security enhancements, as well as quick enterprise support. Tens of thousands of happy customers have a Proxmox subscription. Get yours easily in our online shop.

Buy now!