Error loading state: %[3]s Terraform failed to load the

Error message

Error loading state:
    %[3]s

Terraform failed to load the default state from the %[1]q %[2]s.
State migration cannot occur unless the state can be loaded.
State migration has been aborted. The state in both the
source and the destination remain unmodified. Please resolve the
above error and try again.

What it means

Returned by the single-to-single migration when the source state manager cannot be obtained (`opts.Source.StateMgr(...)` returns error diagnostics). Terraform cannot begin migration without a readable source state, so it aborts with both backends unmodified.

Source

Thrown at internal/command/meta_backend_migrate.go:279

		return fmt.Errorf("Migration aborted by user.")
	}

	// Copy the default state
	opts.sourceWorkspace = currentWorkspace

	// now switch back to the default env so we can acccess the new backend
	m.SetWorkspace(backend.DefaultStateName)

	return m.backendMigrateState_s_s(opts)
}

// Single state to single state, assumed default state name.
func (m *Meta) backendMigrateState_s_s(opts *backendMigrateOpts) error {
	log.Printf("[INFO] backendMigrateState: single-to-single migrating %q workspace to %q workspace", opts.sourceWorkspace, opts.destinationWorkspace)

	sourceState, sDiags := opts.Source.StateMgr(opts.sourceWorkspace)
	if sDiags.HasErrors() {
		return fmt.Errorf(strings.TrimSpace(
			errMigrateSingleLoadDefault), opts.SourceType, sDiags.Err())
	}
	if err := sourceState.RefreshState(); err != nil {
		return fmt.Errorf(strings.TrimSpace(
			errMigrateSingleLoadDefault), opts.SourceType, err)
	}

	// Do not migrate workspaces without state.
	if sourceState.State().Empty() {
		log.Print("[TRACE] backendMigrateState: source workspace has empty state, so nothing to migrate")
		return nil
	}

	srcWord := backendHumanName(opts.Source)
	dstWord := backendHumanName(opts.Destination)

	var err error
	destinationState, sDiags := opts.Destination.StateMgr(opts.destinationWorkspace)

View on GitHub (pinned to d32a084675)

Solutions

Validate the source backend block (`terraform init -backend=false` to isolate, then re-enable).
Confirm the source workspace/key exists and the configured identity can read it.
Re-authenticate and retry `terraform init`.
Inspect the embedded `%[3]s` diagnostic for the precise backend-reported cause.

Defensive patterns

Strategy: validation

Validate before calling

# Validate the source backend can construct a state manager before init:
# confirm bucket/key/workspace resolve and identity has read access.
aws s3 ls "s3://${SRC_BUCKET}/${SRC_KEY}" >/dev/null \
  || echo "source state object unreadable"
terraform init

Try / catch

# Source construction failure is usually config/auth, not transient:
if ! terraform init -input=false >/tmp/init.log 2>&1; then
  grep -q 'Error loading state' /tmp/init.log \
    && { echo 'Source backend config/auth invalid' >&2; exit 2; }
  cat /tmp/init.log; exit 1
fi

Prevention

Keep backend config in VCS and review changes that alter key/workspace paths.
Refresh credentials immediately before init in CI.
Run `terraform init -backend=false` to isolate config-vs-backend issues.

When it happens

Trigger: Fires at the top of `backendMigrateState_s_s` when `opts.Source.StateMgr(opts.sourceWorkspace)` returns diagnostics with errors — invalid workspace name, backend misconfiguration, or auth failure constructing the manager.

Common situations: Backend `key`/`workspace` path is malformed, the source backend's config references a missing bucket/container, or constructing the state client fails due to expired credentials.

Related errors

AI-assisted analysis of hashicorp/terraform@d32a084675 (2026-08-11). Data as JSON: /api/errors/1a2698ecc9781d88. Report an issue: GitHub.