add workaround for azure bug. #5864
Conversation
| // This sleep works around an azure/winrm bug. For more info see | ||
| // https://github.com/hashicorp/packer/issues/5257; we can remove the | ||
| // sleep when the underlying bug has been resolved. | ||
| time.Sleep(1 * time.Second) |
There was a problem hiding this comment.
this seems like the right direction, but my concerns are
- This only impacts ntlmssp connections. should we limit it to those?
- is 1 second long enough? What happens if it's not? Should we make it configurable?
There was a problem hiding this comment.
Is there an easy way to test for ntlmssp connections from inside the provisioner? If so, that's definitely a good idea. And in that case I'd be more willing to bump the sleep length "just in case"; I wanted to keep it short in order to not unduly impact other users. For what it's worth, this actually works on my computer with a sleep of 500ms.
There was a problem hiding this comment.
Is there an easy way to test for ntlmssp connections from inside the provisioner?
probably not, thinking about it...
|
for posterity: this is a hacky workaround to a big hairy bug. We'll continue to investigate in #5865 |
|
I'm going to lock this issue because it has been closed for 30 days ⏳. This helps our maintainers find and focus on the active issues. If you have found a problem that seems similar to this, please open a new issue and complete the issue template so we can capture all the details necessary to investigate further. |
ghost
locked and limited conversation to collaborators
Adds a one-second sleep to the windows-restart provisioner. This works around an azure/winrm bug here: Azure/go-ntlmssp#9 that was causing reliable crashes every time a user ran the windows-restart provisioner on azure.
closes #5257 but doesn't address underlying cause, which is outside of our wheelhouse.