Course contents

SnowPro Core COF-C03 Study GuidePart 5: Data protection and sharingFail-safe

Fail-safe

Fail-safe

Definition

Fail-safe is a fixed, non-configurable 7-day period that begins the instant Time Travel ends for a permanent table. During it, data is recoverable only by Snowflake Support, for disaster scenarios, never self-service.

The full lifecycle of a micro-partition version

Lifecycle of an old micro-partition version, by table type, not to scale. Stages left to right: Active, Time Travel (self-service with AT, BEFORE, UNDROP), Fail-safe (7 days, Snowflake Support only), Purged. Permanent tables: Time Travel 1 day by default, up to 90 days on Enterprise, then 7 days of Fail-safe. Transient tables: 0 or 1 day of Time Travel, no Fail-safe. Temporary tables: 0 or 1 day of Time Travel, ending with the session at the latest, no Fail-safe. Storage is billed for Active, Time Travel and Fail-safe.Activequeries see itTime Travelyou: AT · BEFORE · UNDROPFail-safeSnowflake Support onlyPurgedgone for goodTABLE TYPEPermanentstandard productioncurrentpurgedTABLE TYPETransientreconstructible datacurrentpurgedTABLE TYPETemporarydropped at session endcurrentpurged1 day default · up to 90 *7 days, fixed0 or 1 dayno Fail-safe0 or 1 dayno Fail-safeBilled as storage while the version sits in any of these three stages* Up to 90 days needs Enterprise edition or higher; Standard stops at 1 day.
TABLE TYPEActive
queries see it
Time Travel
you: AT · BEFORE · UNDROP
Fail-safe
Snowflake Support only
Purged
gone for good
Permanent standard productioncurrent1 day default · up to 90 *7 days, fixedpurged
Transient reconstructible datacurrent0 or 1 dayno Fail-safepurged
Temporary dropped at session endcurrent0 or 1 dayno Fail-safepurged

Billed as storage while the version sits in any of these three stages: Active, Time Travel, Fail-safe

* Up to 90 days needs Enterprise edition or higher; Standard stops at 1 day.

Figure 5.2 Every table type starts and ends the same way; what differs is how long Time Travel lasts and whether Fail-safe exists. Read each table type from Active to Purged (stages not to scale): only permanent tables get the 7 Fail-safe days, and only Snowflake Support can reach them.

Fail-safe is a last-resort safety net against catastrophic failure or operator error after Time Travel has lapsed. Every edition has it on permanent tables, and its storage is billed like any other storage. Transient and temporary tables get zero Fail-safe, which is precisely why they cost less to store: Snowflake never has to retain that extra week of old micro-partitions for them.

And it does bill. Fail-safe storage is charged like any other storage for its seven days, which is why a high-churn permanent table costs more than its size suggests: every rewritten micro-partition lingers through Time Travel plus seven more days. Thomas's staging tables rewrite themselves nightly, so as permanent tables with 1 day of Time Travel, each one has nine versions in storage at any moment: one active, one in Time Travel and seven in Fail-safe. Longer retention adds more. Converting them to transient is the standard fix, and it's the answer whenever the goal is to minimize the cost of reconstructible, high-churn data. A recovery that Support runs from Fail-safe also uses serverless compute, which is billed too.

Which tool for which disaster

What went wrongThe recoveryWindow
Bad UPDATE or DELETEQuery or clone with BEFORE (STATEMENT => query_id)Time Travel
TRUNCATE by mistakeINSERT INTO t SELECT * FROM t BEFORE (STATEMENT => '<truncate_id>') (table still exists, nothing to UNDROP)Time Travel
Dropped table, schema, or databaseUNDROP, after freeing the name if something took itTime Travel
Need last week's state alongside today'sZero-copy clone AT a timestamp or offsetTime Travel
Retention expired on a permanent tableSnowflake Support, disaster justificationFail-safe · 7 days
Whole region downAn admin promotes the secondary failover group (see Replication)Business Critical
Table typeTime TravelFail-safeTypical use
Permanent0–90 days7 days, fixedStandard production data
Transient0–1 dayNoneStaging/ETL intermediates, reconstructible data
Temporary0–1 dayNoneSession-scoped scratch work, dropped automatically at session end
Memory aid

"Time Travel you drive, Fail-safe Snowflake drives." You control Time Travel with AT/BEFORE/UNDROP, self-service, instant. Fail-safe has no user-facing controls or syntax at all, it's a Snowflake Support engagement, and it only exists after Time Travel is gone.