Closed new23d closed 5 months ago
@new23d There was a bug in the API which should have been fixed. Does this still happen to you?
@optik-aper this was happening to me earlier today. Has the bug been fixed in the last 4-5 hours?
@new23d Yeah, about 2 hours ago.
Right, I see. Good ol' luck of mine strikes again. I'll close the issue.
I'm glad you noticed it and brought it up any way about it. Thanks for the help.
Is there a place to monitor for API issues at Vultr? I had checked https://status.vultr.com/ and there was nothing related to this.
Yeah, that would be the place. This one was spotted first thing in the AM and resolved within an hour so I don't think a global notification was created this time, but that's where it would go.
Use of the
vultr_instance
resource fails within seconds with{"error":"instance not found","status":404}
.This wasn't happening until 2 months ago I believe. But when I fired up Terraform today, it could not create new instances any more.
I noticed that in the Vultr Web UI, visiting
https://my.vultr.com/subs/?id=45739174-f1ec-46a5-ae4d-038e45e5bc27
(with the UUID copied from the failure output shown below) would complain along the lines of subscription not active yet.Expected behavior
Should be more tolerant to the 404 being returned from Vultr's API when a new instance is created. Maybe the upstream API has changed its behaviour (for the worse.)
Failure Output
Versions Terraform: 1.5.7 Provider: 2.19.0 and master branch
FWIW, I was able to get around by upping the Delay here to 60 seconds and running a local version of the provider. I took the cue from the web UI which in about ~45 seconds would show a legit page for a given instance UUID.