Knowledge base
Everything a search away: what realuptime is, what every error code means and how to fix it, and how to get each product doing useful work in minutes.
How-tos
Create your first monitorFrom signup to a URL being checked from ten regions, in about two minutes.Create a browser checkClick through a real flow in headless Chromium and get a screenshot of what actually rendered.See what came back when a check failedResponse snapshots keep the status line, headers and the first 16 KB of the body from the moment a monitor went down, and from the moment it recovered.Install the monitor agentMonitor hosts and services that the public probes cannot reach, plus server health metrics.Install the agent on LinuxOne line with curl, Docker if you have it, a hardened systemd service if you do not.Install the agent on macOSNode.js 22+ and a launchd job: CPU, memory, disk, load, network, processes and services from a Mac.Install the agent on Windows ServerOne PowerShell line; runs as LOCAL SERVICE under Task Scheduler, no third-party wrapper.Run the agent on KubernetesA DaemonSet that reports every node's health, grouped as one cluster in the dashboard.Set server health alert thresholdsPer host: CPU, memory and disk, each as a line plus how long it must hold, with CPU alerting opt-in.Create a status pageA public, per-region status page for your own customers, on your own domain if you like.Show a vendor's status on your status pagePut Stripe, GitHub or AWS next to your own components, as a labelled third-party reading measured by RealUptime.Show a status banner inside your own appOne script tag that stays invisible until something is wrong on your status page, then shows a dismissible strip linking to it.Translate your status page's chromeShow status words, section headings and the subscribe form in Spanish, German, French, Portuguese, or Japanese; your own content stays exactly as you wrote it.Pause, tag, bulk-edit, and export monitorsStop checking a monitor without losing it, organize a fleet with tags, act on many at once, and download configs or history as CSV or JSON.Get an API keyProgrammatic access to checks, incidents, and status pages over REST and MCP.Restrict an API key to an IP allowlistMake a key usable only from your CI runner, server, or office egress addresses. Requests from anywhere else get a 403.Ask an AI assistant if a service is downConnect realuptime's keyless MCP server to Claude or any other MCP client. No account, no API key, no signup.Set up the errors SDKApplication error tracking: an ingest key, a few lines of SDK setup, and honest drop accounting.Track errors in Rails, Rack and SidekiqThe Ruby gem: one Gemfile line, a DSN, and unhandled exceptions from web requests and background jobs land as issues.Track errors in a Go programThe Go module: a DSN, a client, and net/http middleware or a log/slog hook to catch what your handlers miss.Cap ingest and set an alert thresholdA per-project events/minute ceiling on Growth+, and an absolute occurrence-threshold alert on every plan.Turn on the AI issue summaryAn opt-in, per-project AI-generated summary of an issue's stack trace, breadcrumbs, and occurrence pattern.Install the PHP SDKComposer install, plain PHP capture, and the Laravel service provider: exceptions reported with scrub-by-default and no silent drops.Announce releases from CIAttach a commit and deploy time to an errors release, so the dashboard can show what changed.Register Vercel deploys as releases automaticallyPoint Vercel's own webhook at realuptime and every production deploy shows up as a release, with no CI step to write.Manage checks from a file in your repo (monitoring as code)Declare http checks in realuptime.yaml, plan the diff, apply it from CI, and export what you already have.Install the realuptime CLIHomebrew, or npm from the signed release tarball. Zero dependencies, Node 22 or newer.Sign in with a passkeyAdd a passkey to your realuptime account and sign in with Touch ID, Face ID, Windows Hello, or a security key instead of a password.Upload source maps for JavaScript errorsMinified frames symbolicate only when the release's maps are uploaded: a CLI command, a Vite or webpack plugin, or one curl call.Run status pages for many clients on the MSP planWhat the MSP plan includes, what is pooled across your organization and what is per client account, and how to set it up.Create tickets from Errors issues and incidentsConnect GitHub, GitLab, Jira, Linear, Bitbucket or Azure DevOps once; file one idempotent ticket per issue or incident with a link back.Connect GitHub IssuesA fine-grained personal access token with Issues read and write on one repository.Connect GitLab IssuesA project access token with the api scope, on gitlab.com or a self-hosted GitLab.Connect JiraJira Cloud with an API token and your account email, or Jira Data Center with a personal access token.Connect LinearA personal API key and the team id tickets are filed to.Connect Bitbucket IssuesAn Atlassian API token and account email, on a Bitbucket Cloud repository with Issues turned on.Connect Azure DevOpsA personal access token with Work Items (Read & write) on one organization and project.Open incidents from Alertmanager, Grafana, Datadog or New RelicOne webhook URL per alerting tool opens, updates and resolves incidents on your status page.Install the realuptime Slack appAdd realuptime to a workspace and run /realuptime status or /realuptime incidents from any channel.Check if a vendor is down from Slack or DiscordRun /isdown in Slack or Discord to get RealUptime's own reading on a third-party outage, free, with no account required.Send subscriber email from your own domainPublish a few DNS records and incident, maintenance and confirmation email carries your name instead of RealUptime's, on the Scale plan.Apply for the startup, nonprofit or open-source discount50% off the Growth plan for 12 months, one rule per segment, reviewed by a person.Page on-call through Discord, Teams, Opsgenie, incident.io, FireHydrant, PagerDuty or SlackPoint an escalation step at a notification channel, or let each person choose how they are paged, instead of email only.Get a weekly digest emailOne opt-in email every Monday: uptime and incidents across your monitors, plus new issues, regressions, top issues and dropped-event totals from Errors.Manage monitoring with TerraformChecks, status pages, components and notification channels as code with the realuptime provider, live on the Terraform Registry.Refer a company: give $20, get $20Share your referral link; when the referred account pays its first invoice, both accounts get $20 of credit.Get alerted when a vendor you depend on goes downWatch Stripe, GitHub, AWS or any tracked service in the outage catalog and hear about it the moment RealUptime's own probes confirm an outage, through the same channels your own monitors use.Use your existing Sentry SDKAlready have Sentry's SDK installed? Point its DSN at realuptime and keep it: no SDK swap, no re-instrumentation.Turn on two-factor authenticationAdd an authenticator app, save your recovery codes, and make every sign-in ask for a second factor.Is it down for everyone, or just me? How to tell in 30 secondsA short diagnostic ladder: our outage page for the service, what a difference between the ten regions means, and how to rule out your own DNS, ISP, or VPN before you blame the vendor.What to put on a status page during an incidentA templated, per-region honest-communication playbook: what status to pick, how to start from a real probe alert instead of a blank box, and why our incident text can never say less than the components actually measured.How to monitor an API your business depends on (vendor watch)Watch Stripe, GitHub, AWS, or any of the tracked services in the outage catalog, and get alerted through your own channels the moment RealUptime's probes confirm an outage, not when a crowd starts complaining.Uptime SLAs explained: what 99.9 percent actually allowsThe downtime math behind the common SLA targets, why a vendor's own claimed percentage is not a measurement, and how to check it against an independent probe reading and a permanent incident history instead of trusting the number on the contract.What a heartbeat check is, and when to use one instead of an HTTP monitorCron jobs, backup scripts, and queue workers don't serve requests, so nothing can reach out and check on them. A heartbeat flips the direction: your job pings RealUptime, and silence is what triggers the alert.How to monitor a cron jobCron jobs, backup scripts, and queue workers don't serve requests, so an HTTP monitor has nothing to check. Practical wrappers for a heartbeat monitor instead: where the ping goes in the script, what a grace period buys you, and the two failure stories a heartbeat catches that an HTTP check never sees.Monitor a Coolify app from outside the serverCoolify's own health checks run on the same box as the apps they watch, so the failure that matters most, the server itself going away, is the one they can never report. How to add an independent outside vantage: the Monitor agent on the Coolify host, an HTTP check per deployed app, and a status page.Monitor a Railway app: uptime checks and cron heartbeatsRailway gives you a managed container, not a server, so there is no box to install anything on. What external monitoring looks like on a managed platform: an HTTP check on the public domain, heartbeats for cron jobs and private workers, and error tracking wired to Railway's own deploy variables.External uptime monitoring for a Hetzner server (SysMon is not enough)Hetzner's own tools watch your server from inside Hetzner: console graphs, and SysMon's checks for dedicated boxes, all observe from the provider's own network. What they structurally cannot tell you, and how to add the outside half: probes from independent regions, an agent for the box's own health, and a status page.Monitor Docker containers from outside the hostDocker restart policies and HEALTHCHECK directives live on the same host as the containers they guard, so they die with it. How to watch containers from the outside: an HTTP or TCP check per exposed service, the agent for host and per-container resource usage, and heartbeats for containers that serve nothing.Every refusal realuptime sends carries a stable RU- code: in API response bodies as a code field, and in on-screen messages in parentheses. Each code has its own page at /kb/errors/ru-XXXX with the meaning, common causes, and numbered fix steps. Start at the error-code reference.