all backups stop if a node is offline anywhere in the cluster?!

[Vorl]

Can someone explain this logic to me please? I am completely at a loss for why a company wanting to be an enterprise solution would consider this an acceptable choice to make.

Anyone?

 

[fabian]

this is not the case in general, but without more details, it will be hard to find out what is going on in your particular setup..

 

[Vorl]

you setup a backup, you tell it all nodes.

1 node is offline for any reason none of your backups happen.

 

[echumley]

you setup a backup, you tell it all nodes.

1 node is offline for any reason none of your backups happen.

How many nodes are you running in the cluster? All HA or clustered systems have protections in place to prevent data loss, and Proxmox is no exception, but without more details on your environment, it's hard to tell where your frustrations actually lay.

 

[Vorl]

I have a 5 node cluster, I don't see how having protections in place to prevent data loss translates into "don't backup anything if a node is down"

If anything having a node down means its more critical for backups to run without issues.

Also, what more detail do you need? If you want to see the same thing, take a node down and try to run an all node backup job. It will fail immediately telling you that you have a node down.

 

[UdoB]

First check: pvecm status as you need to have Quorum to start VMs - because a backup does start all "turned-off"-VMs in a "pause"-state. (I am not sure for running VMs...)

1 node is offline for any reason none of your backups happen.

That's not a symptom I can confirm.

You should post the actual error-message you get when you are in that situation.

 

[SteveITS]

I was just going to say post the error, for clarity.

I wonder how many larger-cluster admins have set up one backup job per node, because otherwise all nodes back up in parallel...

 

[fabian]

as a start:
- the "pveversion -v" output of all your nodes
- the storage.cfg
- the backup job config
- the backup task log