What is a lock state?

The lock state in the context of Terraform is a mechanism that prevents multiple operations from being executed simultaneously on the same infrastructure (workspace or backend). It ensures that only one process can modify the infrastructure state at any given time.

This is necessary to prevent issues such as:

• Race conditions: When multiple processes attempt to change the same state simultaneously, which can lead to unpredictable or incorrect results.
• Loss of state data: If one process overwrites changes made by another without knowing about them.
• Inconsistent state: When some changes are applied, and others are not, leaving the infrastructure in an undesirable state.

When a Terraform operation (e.g., terraform apply or terraform destroy) starts, it first attempts to acquire a lock. If the lock already exists, the operation waits or terminates with an error (depending on the configuration). After successfully acquiring the lock, the operation performs its work and then releases the lock.

The location of the lock depends on the backend used:

• Local backend: Uses a local lock file.
• Remote backends (S3, Azure Blob Storage, GCS, etc.): Use the storage's capabilities to organize distributed locking.

An example command to manually unlock (used only in emergency situations):

# Unlock a specific state
terraform force-unlock <lock_id> [DIR]

The lock_id can be obtained from the error message or using the command terraform state list.

Overall, the lock state is a critically important function to ensure integrity and security when working with Terraform in a team environment.