You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
Hi @danielloader , yes its possible for sure. We can introduce a seperate interface for s3-only locking on AWS and gradually deprecate the one based on dynamoDB. There are some open questions:
would the user need to specify the bucket themselves or digger would create a globally unique one on the first run? If the latter than how would digger remember which bucket it used for the locks of that repo - since s3 buckets are globally unique we need a way to remember which bucket was created to store the locks per account. If the user is going to specify the locks bucket then it becomes simpler
Do we keep both dynamodb and S3 as options for AWS locking or deprecate dynamoDB. I think both seem like feasible options to me and could be based on user preference
Uh oh!
There was an error while loading. Please reload this page.
Context
AWS offers strong consistency options on writes on S3 now, so can be utilised as a lock store.
OpenTofu is shipping it next release - opentofu/opentofu#599
Terraform has it enabled already - hashicorp/terraform#35661
The ask
Would it be possible to use Digger with S3 state lock akin to how GCP locking works?
Thanks
The text was updated successfully, but these errors were encountered: