Closed mkulke closed 8 months ago
Hi @mkulke !
the current location (eastus) will only support SEV-SNP and nestable instances, if we want to support TDX runners, we have to move to a region that supports all three of them: westeurope is one of those regions.
Just curious: do we plans to provision bare-metal machines for TDX runners (or sev/snp)?
Since we don't want to invalidate all the urls in the github webhook configurations when moving to another location (garm urls are bound to a location), a discrete
vm_location
param has been introduced that configures GARM to spawn Runners in that region.
Checking I got it right: all the other resources, including the group, are going to still live in eastus?
Just curious: do we plans to provision bare-metal machines for TDX runners (or sev/snp)?
I think we do, but that would not be GARM related, but bare-metal hosts registered as github runners.
Checking I got it right: all the other resources, including the group, are going to still live in eastus?
yup the garm (ACI + storage account etc) resources do still live in eastus.
garm creates a new resource group per runner (with all the suppporting infra, NIC, ip, etc). annoyingly, you cannot configure a location on a per-pool basis, but only globally.
lgtm
the current location (eastus) will only support SEV-SNP and nestable instances, if we want to support TDX runners, we have to move to a region that supports all three of them: westeurope is one of those regions.
Since we don't want to invalidate all the urls in the github webhook configurations when moving to another location (garm urls are bound to a location), a discrete
vm_location
param has been introduced that configures GARM to spawn Runners in that region.