vsphere_virtual_machine timeout waiting for creation creates no tfstate

[matt-deboer]

Often when running terraform apply to create one or more (usually multiple) vsphere_virtual_machine instances, terraform stalls waiting for the vmware tools to come online and report an ip address (quite possibly an esxi problem). This never modifies the .tfstate file, thus subsequent runs of apply will attempt to create the resource again, only to error out finding that it already exists--the net result being manual stop and/or removal of each of the offending vms before running apply again.

It would be better if entries for these vms were added to the .tfstate file in a "tainted" state, finally removing the tainted state once the vm reports back as online. This would allow subsequent runs of apply to delete and recreate the vm as needed, instead of requiring manual deletion, and it would also allow runs of apply to be aborted for stalling out without requiring a bunch of manual intervention.

[jen20]

I think the issue here is whether we get back an ID before the resource creation is complete. If so we may be able to use partial state to record this.

[chrislovecnm]

@jen20 you think this also may be the ip address issue that we are having? @tkak's PR #4283 We are really not waiting for the ip address correctly

[chrislovecnm]

@matt-deboer #6377 may have fixed this ... you mind testing?

[chrislovecnm]

@jen20 @stack72 this should be addressed. Can we close?

[stack72]

Fixed by #6377

[ghost]

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.