I have recently installed a new instance of Cloudstack 4.12 and am reusing a packer job to build templates with that worked in 4.11 but keep getting the error
json: cannot unmarshal string into Go struct field DeployVirtualMachineResponse.ostypeid of type int64
The template_os is described as a uuid and I have checked that it corresponds with the OS type ID in Cloudstack (and tried other OS types). Interestingly, the instance does get created and shows the right OS in Cloudstack after the build has failed.
packer version
= 1.4.0Let me know what further info is required and I will do my best to provide
I bet this falls in one of these two categories:
a) Regression in the CloudStack API
b) Breaking change in the CloudStack API and we need to update the client library we are using. Quickly checking it looks like master of it still uses an int64
for this.
@svanharmelen what do you think?
@rickard-von-essen just a quick FYI that I am working on this one. It's indeed a breaking change in the CloudStack API that requires and update to the client 😉 Hope to be able to provide a fix EOD.
I had a look at this last Friday, but it turns out to be a little more work to sort things out properly. So I'll probably update this issue later this week.
Appreciated, I am a bit stuck at the moment not being able to update my templates.
If this is going to take some time to resolve, is there any workaround that could be used ?
I don't think that is an easy workaround for the time being. But I'm reserving some time for this one early next week. So it shouldn't be too long before we have a fix.
It turns out this takes much more work than expected. Not this specific issue on its own, but for supporting CS 4.12.0.0 as a whole. A lot of tests fail and it seems this is related to either a bug or changed behaviour in CS.
I'm currently discussing my findings with the CS team and hope we can resolve things quickly.
What is the status on this ? What I gather from the associated Cloudstack issue, there will be no fix until next Cloudstack release, is that the end of it or will a workaround be developed ? As you already have established, I can confirm that Terraform is broken as well with the same error.
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.
Most helpful comment
@rickard-von-essen just a quick FYI that I am working on this one. It's indeed a breaking change in the CloudStack API that requires and update to the client 😉 Hope to be able to provide a fix EOD.