[pve-devel] HA Migration on shutdown/reboot

[Bastian Sebode]

Hey Thomas,

thanks for your response.

Am 28.05.2019 um 08:02 schrieb Thomas Lamprecht:
> Hi,
> 
> On 5/28/19 2:40 AM, Bastian Sebode wrote:
>> Hello Proxmox Team,
>>
>> I'm wondering why there is no option "migrate" for the shutdown_policy
>> available.
> 
> Somewhat similar is planned.
> https://bugzilla.proxmox.com/show_bug.cgi?id=2181
> Great. I know it's just nice to have, but it's nice to have! :-D

>>
>> I looked a bit through the code and found create_migrate_worker() in
>> PVE/API2/Nodes.pm, which is used in mass migration. Can't this be used
>> in PVE/HA/LRM.pm when shutdown_policy = "migrate"?
>>
>> You know that servers take from 1 to 10 minutes to reboot. So the
>> interruption by freezing, shutting down and starting on another node in
>> HA setup seems not logic to me and will take a long time to recover the
>> service.
> 
> This _really_ depends on the setup, some even do *not* want live migration
> as they have VMs with hundreds of GB memory, and for them it's much faster
> to shutdown the VM and just restart it over at another node.
> 
Am definitely aware of that, but I bet there are more people who would
like to migrate instead of having downtime. Specially with a fast
migration network. And I'm not proposing that as the default, but as an
option for people with "little" memory.

>> -> shutdown of databases, mounts, guest os
>> -> shutdown of host
>> -> start on other host
>>
>> Migration would keep the service running all the time - I guess in HA
>> activated Environments online migration is mostly possible - and would
>> also accelerate the shutdown of the host, because the VMs don't have to
>> shut down on that host.
>>
>> Defining the Target Node could rely on the HA Groups priority and if not
>> in HA you could still freeze and shutdown a VM or migrate to least used
>> node - I remember Thomas already wrote about this and the problems
>> behind - or even ask where to migrate.
> 
> By changing the HA groups you even can trigger a migration of all VMs,
> but for our planned "maintenance" mode this is not enough, users surely
> want to either migrate, or if local disks, suspend/shutdown non-HA VMs
> as well. And they bring a bit of problems with them, but should be all
> solvable.
> 
Am sure It's solveable but also aware that it takes a lot of brainwork.
Thanks for all that!

>>
>> Right now the only way to keep services in HA online all the time
>> through a node reboot, is to change the HA Group, so the service gets
>> correctly migrated by crm. But that also has to be undone after the reboot.
>>
>> Okay, right now LXC came to my mind... I know there's no online
>> migration, but am speaking for KVM now. ;-) Freeze is also applicable here.
>>
>> Probably it's not that easy to implement as I think on top, but do you
>> already think of a feature like this? Or is there any other way to
>> update my HA activated Cluster without service downtime and without the
>> HA Group changing?
> 
> Yes we think about it, and it's really the next thing on my TODO after
> getting our software stack up and fully ready for the upcoming Buster,
> which is naturally a bit of work for us.
> 
> But your thought-out request is appreciated and I'll try to really
> finally kick-start this.
> 
You're very welcome! And thanks again!!!

> cheers,
> Thomas
> 
Peace & have a nice weekend
Bastian

-- 
Bastian Sebode
Fachinformatiker Systemintegration

LINET Services GmbH | Cyriaksring 10a | 38118 Braunschweig
Tel. 0531-180508-0 | Fax 0531-180508-29 | http://www.linet-services.de

LINET in den sozialen Netzwerken:
www.twitter.com/linetservices | www.facebook.com/linetservices
Wissenswertes aus der IT-Welt: www.linet-services.de/blog/

Geschäftsführung: Timo Springmann, Mirko Savic und Moritz Bunkus
HR B 9170 Amtsgericht Braunschweig

USt-IdNr. DE 259 526 516