This message was deleted.
s
This message was deleted.
h
that sounds like possibly your Typescript code that creates
xxx-policy
might be getting called twice?
t
If that was the case wouldn't the same error occur even on the first deployment? The first deployment works perfectly. If I destroy and redeploy it works perfectly
g
That error looks like it's coming from the provider itself. Are you operating on the correct stack? Or has some other process created a resource with the same name?
t
I always use --stack option, so the correct stack should be picked up by pulumi. No other process or manual intervention creates stacks. BTW, this only happens when I try to update after I have made any changes to the typescript code. I happens specifically with s3 buckets, dynamodb tables, roles and policies. For dynamodb tables still the update seems to carry on, but will roles and policies pulumi throws exception and exits. This is a very undesirable behavior.
Just some additional info, using the following command to update the stack same as creating it
pulumi up --stack <stack> --yes
I have a Pulumi.<stack>.yml file for the specified stack.
So it seems like without auto naming pulumi is unable to update without destroying the stack. Updates should work without auto naming as well and should not throw EntityAlreadyExists exception.
One last problem, when I do an update facing an issue with CloudFront distribution
creating CloudFront Distribution: CNAMEAlreadyExists: One or more of the CNAMEs you provided are already associated with a different resource.
Any help is appreciated
g
Can you share some of the code that is causing this? This is definitely not normal or expected behavior.
t
I apologize for the delay in response, actually I had to move to auto names instead of static names to fix these issues.
101 Views