The Fail-safe period is a critical data protection feature in Snowflake that provides an additional layer of recovery beyond Time Travel. It represents a 7-day period during which Snowflake can recover data that would otherwise be lost, but this recovery can only be performed by Snowflake Support p…The Fail-safe period is a critical data protection feature in Snowflake that provides an additional layer of recovery beyond Time Travel. It represents a 7-day period during which Snowflake can recover data that would otherwise be lost, but this recovery can only be performed by Snowflake Support personnel.
When you delete or modify data in Snowflake, the data lifecycle follows this sequence: Active Data → Time Travel Period → Fail-safe Period → Permanent Deletion. After your configured Time Travel retention period expires (which can range from 0 to 90 days depending on your Snowflake edition and table type), the data enters the Fail-safe period.
Key characteristics of Fail-safe include:
1. **Duration**: The Fail-safe period is fixed at 7 days and cannot be modified or disabled by users. This is a system-level protection mechanism.
2. **Access Restrictions**: Unlike Time Travel, where users can query historical data using AT or BEFORE clauses, Fail-safe data is not accessible through standard SQL commands. Only Snowflake Support can retrieve this data during disaster recovery scenarios.
3. **Storage Costs**: Data in Fail-safe contributes to your overall storage costs. Organizations should factor this into their storage budget calculations, as all data remains billable during this period.
4. **Table Types**: Fail-safe applies to permanent tables. Temporary and transient tables do not have Fail-safe protection, making them suitable for staging data where extended recovery is unnecessary.
5. **Recovery Process**: To recover data from Fail-safe, you must contact Snowflake Support and provide justification. Recovery is performed on a best-effort basis and may take time.
6. **Purpose**: Fail-safe serves as a last resort for catastrophic data loss scenarios, such as accidental bulk deletions or system failures that occur after Time Travel has expired.
Understanding Fail-safe is essential for designing appropriate data retention strategies and calculating total storage requirements in Snowflake environments.
Fail-Safe Period in Snowflake: Complete Guide
What is Fail-Safe?
Fail-safe is a Snowflake data protection feature that provides a 7-day period during which Snowflake can recover data after the Time Travel retention period has expired. This is a last resort recovery mechanism managed entirely by Snowflake and is not accessible to users.
Why is Fail-Safe Important?
• Provides an additional layer of data protection beyond Time Travel • Protects against catastrophic data loss scenarios • Ensures business continuity for critical data • Demonstrates Snowflake's commitment to data durability • Required for compliance in many industries
How Fail-Safe Works
1. Automatic Activation: Fail-safe begins after Time Travel retention expires 2. Duration: Lasts exactly 7 days (non-configurable) 3. Recovery Process: Only Snowflake support can recover data during this period 4. Best Effort: Recovery is performed on a best-effort basis and may take hours to days 5. Storage Costs: You are charged for fail-safe storage
Key Characteristics
• Users cannot query or restore data from fail-safe themselves • Available for permanent tables only (not temporary or transient tables) • Cannot be disabled or modified • Storage is separate from active and Time Travel storage • Recovery requires contacting Snowflake Support
Fail-Safe vs Time Travel Comparison
| Feature | Time Travel | Fail-Safe | |---------|-------------|-----------| | Duration | 0-90 days (configurable) | 7 days (fixed) | | User Access | Yes | No | | Self-Service | Yes | No (Support only) | | Table Types | All permanent tables | Permanent tables only |
Storage Calculation Timeline
For a table with 1-day Time Travel: • Days 1-1: Time Travel storage • Days 2-8: Fail-safe storage • Day 9+: Data permanently deleted
Tables That DO NOT Have Fail-Safe
• Temporary tables • Transient tables • Tables in transient databases or schemas
Exam Tips: Answering Questions on Fail-Safe Period
1. Remember the Number 7: Fail-safe is always 7 days. This value cannot be changed by users, administrators, or account settings.
2. User Access is Restricted: A common exam trap is suggesting users can query fail-safe data. Remember: only Snowflake Support can recover fail-safe data.
3. Table Type Matters: When questions mention temporary or transient tables, remember they have no fail-safe period. This is a frequent exam topic.
4. Sequence Understanding: Data flows from Active → Time Travel → Fail-Safe → Permanent Deletion. Questions often test this order.
5. Cost Implications: Fail-safe storage incurs costs. Questions about minimizing storage costs may involve using transient tables to avoid fail-safe charges.
6. Recovery Time: Fail-safe recovery is not instant. It can take from several hours to multiple days. This is a best-effort service.
7. Common Exam Scenarios: • If asked about recovering data after Time Travel expires → Answer: Contact Snowflake Support (if within 7 days) • If asked about total data retention → Add Time Travel period + 7 days fail-safe • If asked about reducing storage costs → Consider transient tables (no fail-safe)
8. Watch for Trick Questions: Questions may suggest using UNDROP or AT/BEFORE clauses for fail-safe data. These only work during Time Travel, not fail-safe.