Service level agreement
What we commit to for the console and the API, how we measure it from outside, what you get back when we miss it, and the drill numbers it rests on.
Draft: this SLA is not in effect yet. It applies to usage from a date published here; availability before that date is shown for information only. Notes marked “Legal review” are for counsel and disappear when the SLA takes effect.
Commitment
- Availability
- 99.9 % per calendar month
- Allowed downtime
- about 43 minutes in a 30-day month
- Covered
- The console, the public API and the CLI, Terraform and MCP access through it, and the Git, npm and container-registry endpoints the platform serves
- Not covered here
- Your machines, databases, apps, buckets and other resources. They keep running when the console is down; their own service levels are a separate page, not yet published.
- Status page
- status.platform.deutschecloud.app
Legal review: Scope of “covered services” and the definition of a calendar month (time zone).
How downtime is measured
An uptime probe in Helsinki, outside the platform it watches, requests the console’s health endpoint every 30 seconds with a 10-second timeout.
A minute counts as unavailable when the probe gets no successful answer in it. When a second probe location runs, a minute counts only if both see the failure; until then the one probe is binding.
Monthly availability is the share of available minutes in the month. It is the probe’s number, not our server logs.
Your own evidence (times, request IDs) is accepted in a claim. It can add unavailable minutes, never remove them.
Legal review: Whether the probe’s record is binding evidence, and how a dispute is settled.
Service credits
If a month’s availability falls below 99.9 %, your organization receives a credit on its next invoice, as a share of that month’s whole invoice:
| Monthly availability below | Credit |
|---|---|
| 99.9 % | 10 % |
| 99.0 % | 25 % |
| 95.0 % | 50 % |
Credits are capped at 50 % of the month’s invoice and are set against the next invoice. They are not paid out.
Claim within 30 days after the end of the month by writing to info@deutschlandgpt.de, naming your organization and the month.
Free and trial accounts are measured like everyone else and see the same numbers, but receive no credits.
Legal review: Credit base (whole invoice) and the 50 % cap; whether credits are the sole remedy; the 30-day claim deadline; set-off against the next invoice only.
Exclusions
Minutes in these situations do not count as unavailable:
- Scheduled maintenance announced on the status page at least 5 business days ahead, outside 08:00–20:00 CET, at most 2 hours a month. Deploys and planned database switchovers are not maintenance: they are counted.
- Unavailability caused by you: your credentials, your configuration, abuse from your account.
- Force majeure in the narrow sense: war, natural disaster, action by a state authority.
- Features marked as preview.
Counted, not excluded
An outage at our hosting provider, including the loss of a whole data-centre location, and failures of upstream networks or DNS providers count as downtime. Today two of the platform database’s three machines are in Falkenstein, so losing that location stops the console; we count it because it is our risk to remove, not yours to carry.
Legal review: Force-majeure definition and its exclusion of hosting-provider outages; the maintenance notice period and cap.
Recovery targets
For the platform’s own database. These are targets, not credited commitments: they say what our design and drills support.
| Failure | Data loss (RPO) | Back in service (RTO) |
|---|---|---|
| One database machine lost | none | ≤ 1 minute |
| The whole database cluster lost | ≤ 5 minutes | ≤ 1 hour |
Legal review: The “targets, not credited” clause.
What the drills measured
On 2026-11-13, on a copy of the production layout (a load balancer, two control-plane machines in Falkenstein and Nuremberg, a three-machine PostgreSQL cluster), measured from outside with a request every 250 milliseconds and a write every 500 milliseconds:
| Failure | What clients saw | Writes lost |
|---|---|---|
| A control-plane machine powered off | no failed request; some answers took about 5 s for 8 s | 0 |
| The app on one machine stopped | about half the requests failed for 9 s | 0 |
| The database leader’s machine powered off | every request failed for 30 s (44 s with the database’s default failover timing) | 0 |
| Planned database switchover | every request failed for 3 s | 0 |
| All three database machines lost, restored from backup | 2.5 minutes, with the steps prepared | the last 4.6 minutes |
| The Falkenstein location lost | down for as long as it was, plus 40 s | 0 |
Measured availability
Published here for every month from the date this SLA takes effect.
Start building on Deutsche Cloud
Create an account with your company details, install the command line and deploy your first app. Pay only for what runs.