blog/sqliaas-problem/ about blog HOT 1 OPEN

Hey Mark - I have this issue with using terraform to provision a VM SQL server. I've found if I let terraform try to provision it, it'll fail with the "Execution timeout" error you listed above. However, if I delete the SQL VM with this command: az sql vm delete --name (name) --resource-group (name --yes and then re-run terraform, it works immediately.

My suspicion is this is closely related - potentially deleting the SQL VM restarts the SQL service on the host VM and the secondary deploy succeeds?

I'm playing around with timers to see if it's related to load or initial VM boot, but it doesn't appear to be - I can add a delay to the provisioning of 5m (I'm testing longer delays), and it doesn't help. And the VM's initial CPU load is pretty small - 5-10% during this initial time, it's certainly not pegged. I'm deploying to relatively beefy machines (Standard_E20s_v5), so I don't think resource constraints are what's happening here.

I haven't yet RDP'd into the host to validate the SQL logs and if they indicate any particular issue.

Did $MSFT ever get back to you with more information? Thanks so much for posting regardless.
Kyler

from blog.