Uninstall FormKiQ
Uninstalling removes infrastructure and can permanently destroy documents, identities, configuration, and audit evidence. Start by deciding what must be preserved and who has approved deletion. These instructions describe an operator procedure; they do not require deleting a running environment merely to verify installation.
Inventory and preserve
- Identify the exact AWS account, region, root stack, nested stacks, optional support stacks, and integrations. Record outputs and parameters before removal.
- Stop new uploads, scheduled jobs, webhook producers, and external synchronization. Plan how to prevent late deliveries or automatic reprovisioning.
- Export required documents, versions/artifacts, metadata, configuration, identities/group assignments, audit records, templates, and keys under your recovery policy. Confirm a usable restore in a separate environment.
- Check native legal holds, retention requirements, storage Object Lock, backup retention, and contractual obligations before approving deletion. CloudFormation deletion is not a legal hold override procedure.
- Record shared infrastructure. A VPC, certificate, identity provider, notification configuration, or search service used by another application must not be deleted as part of this stack's cleanup.
See Backup and Recovery and Retention and Legal Hold.
Remove the root stack
Inspect CloudFormation termination protection and the template's EnableDeletionProtection settings. They are different controls. If deletion is authorized, change the relevant protections through the owning stack/configuration after preserving required data.
Delete the root application stack so CloudFormation coordinates its nested stacks. Do not start by deleting individual child resources. The following is an example command to run only after the account/stack and data-retention decision have been verified:
aws cloudformation delete-stack \
--stack-name YOUR_APPROVED_STACK \
--region YOUR_REGION \
--profile YOUR_PROFILE
aws cloudformation wait stack-delete-complete \
--stack-name YOUR_APPROVED_STACK \
--region YOUR_REGION \
--profile YOUR_PROFILE
If deletion fails, inspect stack events. Common causes include nonempty/versioned buckets, DynamoDB deletion protection, dependencies, protected storage, or permissions. Do not issue blanket account-wide bucket or table deletion commands to make the stack pass.
S3 versioned buckets can contain old versions, delete markers, and multipart uploads that an ordinary object listing does not show. Retained resources and snapshots can remain billable after the stack is removed. Review the exact released template and your deployed retention policies; do not assume either that every resource is retained or that stack deletion erases all data.
Remove optional infrastructure and external access
After checking dependencies, remove only the dedicated support stacks that are no longer needed: search, bastion, VPC, backup, certificate, or other module infrastructure. CloudFront distributions, DNS records, certificates, KMS keys, secrets, logs, and backups may require separate lifecycle handling.
Revoke integration keys, disable external webhook targets, update SSO applications, remove obsolete CI deployment roles, and disconnect signing/AI/email integrations where appropriate. Remove browser/client configuration pointing to the retired deployment. Respect key-deletion waiting periods and backup-retention requirements.
Verify completion
Confirm the root stack's deletion outcome, inventory retained resources by ownership tags and recorded IDs, check billing/resource reports, and test that retired API/console hostnames and integrations no longer provide unintended access. Record retained data, its owner, access controls, retention deadline, and eventual disposal procedure.
A successful stack deletion is not evidence that external exports, provider records, or backup copies have been erased. Keep the final inventory and approval evidence with the operational record.