Snowflake Postgres (HA instances) SLASnowflake · Databases · Snowflake · Databases
Promise vs reality
Downtime used against the allowance
5.3 d of major incidents, 28.9x the allowance4.4 h allowed in 365 days
30 major and 33 minor incidents; worst region Azure - East US 2 (Virginia). 29 h degraded (not counted).
Major incident minutes per month
Allowed per month: 22 min
- Promised
- 99.95%
- Observed, 365d
- 98.557%
- Major incident time
- 5.3 d
- Credit on first breach
- n/a
Commitments and credits
- High-Availability instanceSnowflake Postgres instances with the High-Availability deployment option; non-HA and burstable instances are not covered99.95%
Monthly uptime below, credit of the bill
No credit schedule published for this tier.
Terms
- Measured
- Monthly HA Instance Availability Percentage per HA Instance; downtime = one-minute intervals in which all connection requests fail
- Credit cap
- Total HA Instance Credits for the month (i.e., the full month's consumption of that instance) at the worst tier
- Claim window
- Within twenty-one (21) days of the calendar month in which the failure occurred
- How to claim
- Customer requests Feature Level Credits; credited against usage of the affected HA Instance the following month
Not covered
- All Section III Support Policy exclusions
- Burstable instances
- Resource exhaustion (CPU, memory, IOPS, storage) from customer workloads
- Customer-initiated restarts or misconfigurations
- Postgres engine software causing repeated crashes
- Scheduled maintenance initiated by customer or Snowflake
Less than 99.95% but equal to or greater than 99.0% 3x Average Daily HA Instance Credits; Less than 99.0% but equal to or greater than 95.0% 7x ...; Less than 95.0% Total HA Instance Credits
Credits expressed as multiples of Average Daily HA Instance Credits (3x, 7x, total), so numeric credit is null. Added in the February 24, 2026 version (absent from the January 9, 2026 version).
Evidence: incidents, last 365 days
- Sep 14, 2026INC20000217AWS - Asia Pacific (Sydney), AWS - Asia Pacific (Singapore), Azure - Australia East (New South Wales), AWS - US East (Ohio), AWS - Asia Pacific (Tokyo), AWS - Asia Pacific (Jakarta)Degraded
- Sep 11, 2026INC20000213AWS - Europe (Frankfurt), AWS - Asia Pacific (Tokyo)Outage
- Sep 9, 2026INC20000211AWS - US East (N. Virginia)Outage
- Sep 4, 2026INC20000199AWS - Europe (Frankfurt)Outage
- Sep 1, 2026INC20000190AWS - US West (Oregon), AWS - GovCloud (US-West), AWS - Canada (Central), AWS GovCloud - DoD, AWS - Europe (Ireland), AWS - Europe (London)Degraded
- Aug 31, 2026INC20000188AWS - US East (N. Virginia)Outage
- Aug 28, 2026INC20000150GCP - US Central 1 (Iowa)Minor
- Aug 27, 2026INC20000182Azure - Central US (Iowa), Azure - Central India (Pune), Azure - Australia East (New South Wales)Degraded
- Aug 25, 2026INC20000175Azure - Southeast Asia (Singapore)Outage
- Aug 24, 2026INC20000173AWS - US West (Oregon)Degraded
- Aug 20, 2026INC20000168Azure - East US 2 (Virginia)Degraded
- Aug 18, 2026INC20000163AWS - US East (N. Virginia)Degraded
- Aug 14, 2026INC20000158Azure - East US 2 (Virginia)Outage
- Aug 14, 2026INC20000157AWS - US East (Ohio)Outage
- Aug 5, 2026INC20000139AWS - US East (N. Virginia)Minor
- Aug 3, 2026INC20000129AWS - US East (Ohio)Outage
- Aug 3, 2026INC20000128AWS - Asia Pacific (Tokyo)Minor
- Jul 30, 2026INC20000115AWS - US West (Oregon), AWS - Canada (Central), AWS - US East (Ohio), AWS - Europe (London), AWS - Europe (Frankfurt), AWS - Asia Pacific (Mumbai)Outage
- Jul 25, 2026INC20000106Azure - Central US (Iowa)Minor
- Jul 11, 2026INC20000085AWS - US West (Oregon)Minor
- Jul 10, 2026INC20000082Azure - East US 2 (Virginia)Degraded
- Jun 30, 2026INC20000063AWS - US West (Oregon)Outage
- Jun 24, 2026INC20000052AWS - US East (N. Virginia)Outage
- Jun 23, 2026INC20000047AWS - US West (Oregon)Degraded
- Jun 23, 2026INC20000046AWS - Europe (London)Degraded
SLA changes
- Feb24Snowflake Postgres (HA instances)Appendix A added a Snowflake Postgres HA Feature Level of 99.95% with 3x/7x/total HA-instance credits; not present in the January 9, 2026 version.Source
Other databases SLAs
ServiceDowntime vs allowed
Amazon DynamoDBAWS99.999% / 99.178%
Azure SQL DatabaseMicrosoft Azure99.995% / n/a
Lakebase PostgresDatabricks99.99% / n/a
Cloud SQLGoogle Cloud99.99% / 99.927%
Managed DatabasesDigitalOcean99.95% / 100%
Elastic Cloud Hosted (formerly Elasticsearch Service)Elastic99.95% / 99.146%
Snowflake Postgres (HA instances)Snowflake99.95% / 98.557%
Promised in the SLAObserved on the vendor's status pagePartial historyShortfall
Summaries of published SLAs; the contract you sign governs. Logos via logo.dev; trademarks belong to their owners.