Skip to content
cloudhost
Observability

Know what your app does before a user tells you.

Logs, metrics and a trace of each request for every app and function, with nothing to install. Set alert rules and hear about it by email, SMS, Bale, Eitaa or a webhook when something breaks.

What you get

  • Logs

    Every request, console line, cron run and outgoing fetch. Filter by level, status, path, source, environment, branch or version, and search the text, with a chart over time and counts per filter.

  • Live tail

    Follow an app's or function's log as it is written, in the dashboard or the terminal.

  • Metrics

    Requests, 4xx and 5xx errors, the error rate and p50, p95 and p99 latency over time, per project and per route. Latency is measured from request to response, cold starts included.

  • Metrics for every product

    SQL databases, caches, buckets and queues each have their own Metrics tab. For functions, see how many copies ran on each server and whether any request waited.

  • Traces

    Every request has its own trace id. Calls to the SQL database, cache, storage and fetch become spans on their own, and you see the request's path as a waterfall. Values, keys and bodies are never recorded.

  • Traces that join your other systems

    The standard traceparent header is accepted and handed to your code. Pass it on in your own fetch calls to join your other services' traces. The response carries the trace id too.

  • Alerts on all three kinds of data

    Alert rules on metrics, logs and traces: when a value stays past a limit for a while, changes suddenly, leaves its usual pattern, or one member of a group behaves unlike the rest.

  • Alerts for what should or should not happen

    When data you expect stops arriving, a value is seen for the first time, a trend will cross a limit within a set time, one event follows another, or an address is on your known-bad list.

  • Alerts wherever you are

    Email, SMS, a webhook, a Bale or Eitaa bot, or a notification on your phone. Each rule tells you when it starts firing and when it resolves. Try the channels with a test send.

  • Log export

    Send a project's or the whole team's logs to your own tools over HTTPS or syslog over TLS, signed, with source and level filters and sampling. If the endpoint goes down, sends retry with backoff and nothing is skipped.

  • Webhooks and events

    Send deploys, builds, alerts and other events to your own service as signed webhooks, with a log of every delivery and redelivery.

  • Activity log

    Who changed what, and when: every action by team members and API keys is in the team's activity log.

From the terminal, the dashboard or the API

The same logs, traces, alerts and exports in the dashboard, the CLI, the API and MCP. Your coding assistant can read logs and find the error too.

cloudhost logs my-api --project shop --level error --status 5xx --since 1h
cloudhost logs my-api --project shop --search "payment failed" --environment production
cloudhost tail my-api --project shop

Included with every project

Logs, metrics, traces and alerts come with every app and function and are not metered on their own. Logs and traces are kept for a set time, the same for every team. Log export is on Pro. The details are on the pricing page.

Questions

Deploy your first app and watch its logs

From the first request, with nothing to install.

Sign-up is by invite for now. If you have a code, create your account.

Sign up withEmail codeSMS code