Databricks SLA
99.99% uptime guarantee (Lakebase High Availability endpoints on paid Platform Services plan only)
Official SLA Commitment
Applies to
Lakebase High Availability endpoints on paid Platform Services plan only
Scope
Lakebase Postgres endpoints configured for High Availability on Databricks Platform Services
Measurement
Monthly Uptime Percentage = (total calendar-month minutes − Downtime minutes) ÷ total calendar-month minutes × 100; Downtime confirmed per Lakebase SLA definitions
Exclusions
Force majeure, customer/third-party equipment or connectivity, customer misuse or Agreement violations, scheduled maintenance per SLA exclusions
Not covered
General multi-cloud Platform Services (MCSA disclaims uninterruptibility); free trials; Beta services; customer-managed compute and storage; Private Cloud deployments without separate SLA
Promised vs Measured
SLA Promise
99.99%
≤ 4m downtime/month
Measured (30d)
—
Collecting data…
Verdict
—
24h Uptime
—
7d Uptime
—
30d Uptime
—
Pricing & SLA by Tier
| Plan | SLA Guarantee | Price | Meeting 30d? |
|---|---|---|---|
| Free trial | No uptime SLA | $0 (14-day trial; cloud infra costs may apply) | — |
| Premium (pay-as-you-go) | No uptime SLA | From $0.22/DBU (SQL Classic); usage billed per second | — |
| Committed Use Contracts | No uptime SLA | Custom discounts on committed DBU/DSU spend | — |
| Enterprise | No uptime SLA | Custom (contact sales; ~15–25% above Premium per industry reports) | — |
Source: databricks.com/pricing
SLA Credit Policy
Service credits vary by sub-service. Lakebase credits are a percentage of affected endpoint compute fees; Azure Databricks credits are a percentage of monthly workspace fees per Microsoft SLA.
| Monthly Uptime | Credit |
|---|---|
| Lakebase HA: <99.99% and ≥99.9% | 5% of monthly fee |
| Lakebase: <99.9% and ≥99.0% | 10% of monthly fee |
| Lakebase: <99.0% and ≥95.0% | 25% of monthly fee |
| Lakebase: <95.0% | 50% of monthly fee |
| Azure Databricks: <99.9% | 10% of monthly fee |
| Azure Databricks: <99.0% | 25% of monthly fee |
Credits capped at 50% of monthly fee. Applied within 45 days of approval.
Industry SLA Tier Compliance
| Tier | Downtime Budget/Month | 24h | 7d | 30d |
|---|---|---|---|---|
| 99.9%(three nines) | 43m 12s | |||
| 99.95% | 21m 36s | |||
| 99.99%(four nines)Databricks's SLA | 4m 19s | |||
| 99.999%(five nines) | 26s |
Equivalent Downtime
Last 24 hours
—
Last 7 days
—
Last 30 days
—
Incident Summary (90 days)
Total Incidents
13
Active Now
0
Resolved
13
Avg. Resolution
2h 23m
Databricks SLA: What You Need to Know
Databricks guarantees 99.99% uptime measured as lakebase postgres endpoints configured for high availability on databricks platform services. This commitment applies exclusively to Lakebase High Availability endpoints on paid Platform Services plan customers.
Based on DevHelm's continuous monitoring, Databricks currently shows a 30-day measured uptime of —. Uptime data is still being collected for a complete assessment.
What's Covered (and What Isn't)
Databricks's SLA specifically covers lakebase postgres endpoints configured for high availability on databricks platform services. It explicitly excludes: General multi-cloud Platform Services (MCSA disclaims uninterruptibility); free trials; Beta services; customer-managed compute and storage; Private Cloud deployments without separate SLA. This means outages in excluded components don't count toward SLA credits.
Impact on Your Application
Once sufficient monitoring data is collected, this section will show projected annual downtime. If your application depends on additional services, their independent failure probabilities compound — making dependency monitoring essential for maintaining your own SLA commitments.
Monitoring Methodology
DevHelm tracks Databricks's official status page continuously and aggregates component-level status data into uptime percentages. Data is refreshed every 60 seconds. Uptime is calculated as the percentage of time all components report operational status.
