-
Notifications
You must be signed in to change notification settings - Fork 1.7k
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
Cannot start a rolling update on an unhealthy Managed Instance Group #10648
Cannot start a rolling update on an unhealthy Managed Instance Group #10648
Comments
@ScottSuarez I'm tagging you in this issue since you resolved the last issue and probably have the most context already. Thanks for taking a look! |
This is what wait for instances` does and how terraform works. Unfortunately there is very little we can do here. The ask is contradictory. You want us to wait for the status. If terraform never acquires the desired status then we need to fail. If we allow it to pass then the field is not actuating anything and just delaying deployment. I understand how this can be difficult but I would suggest not uses wait for instances and monitoring the status manually. To allow this to pass would contradict terraform's deployment methology |
@ScottSuarez I gotcha, that makes more sense with how you are viewing I think the issue is that even during the I don't believe this expectation to be contradictory and still seems very aligned with how terraform operates and the normal workflow. Does that make more sense? |
yes that makes sense to me, I've put out a fix for your scenario. |
Thank you so much! |
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. |
Community Note
modular-magician
user, it is either in the process of being autogenerated, or is planned to be autogenerated soon. If an issue is assigned to a user, that user is claiming responsibility for the issue. If an issue is assigned tohashibot
, a community member has claimed the issue already.Terraform Version
Terraform v1.0.11
Affected Resource(s)
Terraform Configuration Files
Please see this issue for example code.
Debug Output
Panic Output
n/a
Expected Behavior
I should be able to make code changes to start a new rolling update when the managed instance group has an existing operation in progress.
Use Case 1
I need to make configuration updates to my instance template which will require a rolling update; however, I initially make a mistake and the rolling update stalls indefinitely while waiting for the new node(s) to pass health checks. The MIG continues to auto-heal according to the policy, trying to get the MIG back into a healthy state. According to the Google docs for rolling updates, a rollback is performed by starting another rolling update. In order to fix the MIG, I need to be able to make the changes in terraform and I expect terraform to be able to start the rolling update.
Use Case 2
I have a deployed MIG in production, but one of my nodes starts failing, putting the MIG in an unhealthy state. The failure is caused by something external to the MIG which prevents auto-healing from resolving the issue as new nodes also fail. I need to be able to make code changes through terraform and start a rolling update to the unhealthy MIG.
In both of these use-cases, terraform should be able to start a new rolling update regardless of the current state of the MIG, just like an operator would do manually with
gcloud compute instance-groups managed rolling-action start-update
.Actual Behavior
Terraform times out during the
apply
phase, waiting for the MIG to become healthy.Steps to Reproduce
Important Factoids
References
Please review the history in this issue that discusses the previous problem with even doing a
plan
against an unhealthy MIG.The text was updated successfully, but these errors were encountered: