Your telemetry. Our pager.
TracePath Cloud is the hosted version of the whole product: logs, traces, metrics, session replay, exceptions, profiles, AI tracing, synthetic monitors, status pages and on-call paging. You send OTLP. We keep the thing running.
No card on the Free plan. Two projects, 5 GB a month, no expiry.
The parts nobody wants to own.
Running an observability stack is its own on-call rotation: storage that outgrows the disk, collectors that fall behind, upgrades nobody scheduled. That rotation is ours.
Upgrades and migrations
New versions and schema migrations are ours to run. You find out because a feature appeared, not because a maintenance window did.
Storage that keeps up
Columnar storage sized and tuned for the shape of telemetry, so a bad afternoon in production does not turn into a query timeout.
Teams, roles and on-call
Organisations, per-project role overrides, escalation policies, schedules and paging to email, Slack, Telegram, Pushover, SMS or a webhook.
Ingest that just takes OTLP
Point any OpenTelemetry exporter at the ingest endpoint with a project token. No agent of ours to install, no proprietary wire format to adopt.
Published prices, published limits.
Four plans, listed in euro, with the limits written down rather than negotiated one customer at a time. Ingest pauses at the allowance instead of turning into an invoice.
Prices exclude VAT. EU business customers with a valid VAT ID are invoiced under the reverse charge; Polish customers are charged 23% VAT.
What we can actually claim.
Certification logos are easy to put on a page and hard to earn. Here is the list of controls that exist today, and the honest note about the ones that do not.
Hosted in the EU
TracePath Cloud runs on infrastructure in the European Union, currently at Hetzner in Helsinki, Finland. Every provider we rely on is named on the subprocessors page.
Encrypted in transit, hashed at rest
TLS on every endpoint, ingest included. Passwords are hashed, and personal access tokens, refresh tokens and SMS verification codes are stored only as hashes, so a database read does not hand anyone a working credential.
Access you can actually scope
Four organisation roles plus per-project overrides, revocable personal access tokens, OAuth device flow for the CLI, and per-project ingest tokens you can rotate.
Paperwork on request
The data processing agreement is part of the Terms on every plan, and a countersigned copy is available on request. We do not hold a SOC 2 or ISO 27001 report, we do not offer a BAA today, and we will not imply otherwise.
More detail on security, subprocessors and the data processing agreement.
Standard protocol. Standard exit.
Your applications are instrumented with OpenTelemetry, not with us. Everything arrives over plain OTLP, which every other OpenTelemetry-compatible backend also speaks.
- Instrument once with the OpenTelemetry SDK for your language, then point the exporter wherever you want it.
- Moving to or from TracePath is an endpoint change and a token change, not a re-instrumentation project.
- Dashboards export and import as JSON, so the views your team built travel with them.
- Your stored telemetry is readable through the documented query API with a personal access token. There is no one-click bulk export yet: ask support and we will produce an export within 30 days.
Detect. Replay. Resolve.
Create an account, point an OTLP exporter at us, and watch the first trace land.
Cloud FAQ
What do I actually get on TracePath Cloud?
The whole product, on every plan. Logs, distributed traces, metrics and dashboards, exception tracking with automatic grouping and ranking, web session replay attached to the exception that produced it, continuous profiling, AI and LLM tracing with cost and token accounting, synthetic monitors with public status pages, incidents and post-mortems, and on-call schedules with escalation policies. Plans change the volume, the retention window and the number of people and projects, not the feature list.
How do I send data?
Over OTLP/HTTP. Instrument your service with the OpenTelemetry SDK for its language, set the exporter endpoint to the TracePath ingest host and pass your project token as a bearer credential. Protobuf and JSON payloads are both accepted, gzip optional; there is no gRPC endpoint. Browser and mobile applications can use the OpenTelemetry web and mobile SDKs today; the first-party TracePath SDKs that add session replay and source-map symbolication are in private beta — email support for access.
How is this different from running it myself?
You skip the storage engine, the upgrades, the retention workers and the capacity planning, and you skip the point six months in where the telemetry store becomes its own production system with its own on-call rotation. What you give up is control over the hosts. Because everything arrives over standard OpenTelemetry, your instrumentation is never locked to us.
Where does my data live, and who can touch it?
On infrastructure in the European Union, currently at Hetzner in Helsinki, Finland. Access inside TracePath is limited to the people who operate the service, and every third party involved in delivering it is listed on the subprocessors page before we start using them.
Can I move off TracePath later?
Yes, and the instrumentation is the reason it is cheap. Standard OTLP means switching backends is a configuration change rather than a rewrite. Dashboards export as JSON, and your stored telemetry is readable through the query API with a personal access token. There is no one-click bulk export yet — ask support and we will produce an export within 30 days, while your account is active.
What happens if I send more than my plan allows?
At 80% of the monthly ingest allowance the organisation owner gets an email. At 100% ingest pauses with an HTTP 503 and a Retry-After header, which OTLP exporters and the TracePath SDKs treat as a retry rather than a failure, and the rest of the product keeps working on what is already stored. It resumes next period or on upgrade. There is no overage rate.
Do you support single sign-on?
Not yet. Accounts sign in with email and password, protected by email verification and optional two-factor authentication, or with a Google or GitHub account. Per-organisation SSO is on the roadmap; tell us if it is a requirement for you.
How do I get help?
Email [email protected]. You reach the people who build TracePath, not a ticket queue in front of them. Business and Enterprise customers are answered first, and Enterprise customers get a named contact and agreed response targets.