Terraform

Problem

`terraform plan` or `terraform apply` fails immediately with an error that the state is locked.

Symptoms

• Error acquiring the state lock
• Lock Info block showing ID, Path, Operation, Who, Version, Created
• Started after a previous run was killed (Ctrl+C, CI job canceled, machine crash)

Root Cause

Terraform uses a lock (a DynamoDB row for the S3 backend, or a native lock for Terraform Cloud) so two runs can't write state at once. A run that didn't exit cleanly leaves that lock held even though nothing is actually running anymore.

Solution

terraform plan
# note the Lock ID in the error output

terraform force-unlock <LOCK_ID>

aws dynamodb scan --table-name <lock-table-name>

terraform plan

Prevention

• Never kill a running `terraform apply` mid-write — let it finish or fail on its own
• Use a CI pipeline with concurrency limits so two applies against the same state can't start at once
• Keep state locking enabled (the default) rather than disabling it to work around this
• Use short-lived CI runners so a crashed job's lock naturally clears

References

• Terraform State Locking
• force-unlock command