Scenario: you disable your key at 2 AM
What happens immediately
When you disable a key version, GCP stops honoring encrypt/decrypt requests for that version within seconds. Here’s the cascade:| Time | What happens |
|---|---|
| T+0s | Key disabled in GCP KMS |
| T+5–30s | In-flight Parable API requests that require decryption start returning 500 errors |
| T+60s | Parable’s background services detect the key failure and stop retrying |
| T+2–5m | Connector ingestion jobs fail with KMS_KEY_DISABLED error |
| T+5m | Parable on-call receives an alert about key failures (we monitor decryption errors) |
Parable will not re-enable your key without authorization
Parable’s application service accounts can only use keys (encrypt/decrypt), not manage them. Parable’s infrastructure administrators have broader access for operational purposes, but will never re-enable your key without your explicit written authorization.Recovery: re-enabling the key
Sign in to GCP Console
Sign in using your federated sign-in link or Google account (see Key Management for sign-in details), then navigate to your keyring: Security → Key Management →
tenant-{your-slug}-kms.Re-enable the key version
Find the disabled version, click the three-dot menu → Enable.Confirm in the dialog. The status changes to Enabled immediately.
Services recover automatically
Parable’s services retry failed operations and resume normal function within 60 seconds — no action needed on your end and no need to contact support unless issues persist after 5 minutes.
Checklist: disabling your key
Use this checklist before disabling your key in a security incident:- Notify your Parable account manager or email support@askparable.com — we can help triage whether KMS disable is the right response
- Note the current time — you’ll need it for audit log review later
- Disable the key in GCP Console (see Key Management)
- Confirm Parable services show errors (expected) — verify at least one API call fails
- Document the incident and gather audit logs for the period before the disable
- When the incident is resolved, re-enable the key (steps above)
- Trigger manual syncs for any connectors that failed during the outage