How can I cleanup old graphs for a VM id?

[elicro]

I am testing VE and I noticed that even after deleting the VM the graphs are still present on the system.
How can I see them? just create a new vm with the same ID ie 100, remove, spin a new VM.
Then when you access the VM you see that the graphs are still there with the data of the old VM and the data is being updated with the new VM.
It's OK as its but I need to be able to purge the graphs, any recommendation on how to do that?

Thanks!

 

[Dunuin]

First, you shouldn't use a VMID twice. If you got a VM with VMID 100 and you destroy it and create another VM, use the next free VMID that was never in use. Otherwise you will run into problems when for example using PBS and it might get very confusing with archiving VZDump backups and documentation. So you will have to manually keep track of all VMIDs by doing proper documentation. And in case you got multiple nodes, make sure no VMID is ever used twice across all those nodes.

If you still want to clear the graphs, those are stored via rrdcache. But not sure how to clean that for a single VM.

 

[elicro]

@Dunuin If the VMID is so "sensitive" like what you say then I believe it should be unique and should be used as Counter with Auto Increment option ie can be customized but usually will be automatic.

I will test couple things later.
Thanks!

 

[Dunuin]

@Dunuin If the VMID is so "sensitive" like what you say then I believe it should be unique and should be used as Counter with Auto Increment option ie can be customized but usually will be automatic.

There are enough discussions and feature requests about having some kind of auto increment because people don't like to manually keep track of not using a VMID twice...
See for example here: https://bugzilla.proxmox.com/show_bug.cgi?id=4369