terraform-provider-iterative
terraform-provider-iterative copied to clipboard
iterative
☁️ Terraform plugin for machine learning workloads: spot instance recovery & auto-termination | AWS, GCP, Azure, Kubernetes
This is could be applicable for the other providers but using `gcp` I received a `"Quota 'FIREWALLS' exceeded. Limit: 100.0 globally."` Should we query the API for an existing iterative...
Coming from [discord](https://discord.com/channels/485586884165107732/728693131557732403/986933583832813648) ```sh NVIDIA-SMI has failed because it couldn't communicate with the NVIDIA driver. Make sure that the latest NVIDIA driver is installed and running. ``` I have confirmed...
Finding topics within the terraform docs is far from ideal. - can we hack it? - if not... should we move from there? I have the same questions for Discord...
Collection of data handling issues ## new features - [x] #360 - [x] #299 + prerequisite for migrating runner to task? - [x] #590 - [ ] #289 ## improving...
With the current implementation, instances can't destroy all the supporting resources, because of interdependency. For example, after deleting a security group, it's impossible to issue more API calls because there...
It would be great for `workdir`/`output` to support: * Lists of multiple paths * Globs or regexes
The current mechanism used to pass environment variables to `task` scripts stores the values in plain text as part of the instance's `cloud-init` startup script. Instead, it should be using...
> _the auto-scaling group doesn't get the tag, and more importantly, s3 objects are missing them as well_ * https://github.com/iterative/terraform-provider-iterative/pull/599#issuecomment-1144964869
_**(Continuation of #533)**_ Once the quota increase request is approved, we should use `Standard_NC4as_T4_v3` instead of `Standard_NC6` to run the tests: https://github.com/iterative/terraform-provider-iterative/blob/765225c9b967a25cbdb8719370b3899665e024b4/.github/workflows/smoke.yml#L100
### Follow-up of #420 and #453, closes #459 It turns out that #459 happened because when some resources are already missing, `t.Read()` fails and the emptying logic never runs. Still,...