Skip to main content
Back to Solutions

You find out before your users do, without waking up for blips

Uptime checks with retries and cross-location verification, plus SSL, DNS and domain checks for every subdomain, in one workspace. Incidents go to Slack or to the on-call tool you already use.

No credit card required·Verification and incidents on paid plans

Uptime

A health route returns 200 while the database is down

When the first sign of an outage is a support ticket, the incident starts already behind. An uptime monitor checks each endpoint from multiple global locations, such as London and New York, on an interval you set per monitor, from 30 seconds to 24 hours depending on your plan.

A health route can return 200 while its body says the database is gone. A keyword that must or must not appear catches that, and you decide which status codes count as up. Monitors can send Basic, Token or Digest auth, custom headers and a request body, so protected API routes get the same coverage as public pages.

Every check records its location, response time, status code and response headers, and the response-time chart helps you spot slowdowns before they become outages.

Uptime monitors

Production

  • api.example.com/health

    GET · token auth · 200, body contains "error"

    188 ms

    Down
  • app.example.com/login

    POST · JSON body · accepts 200, 302

    318 ms

    Up
  • docs.example.com

    GET · must contain "Getting started"

    96 ms

    Up
  • billing.example.com:8443/status

    GET · timeout 10 s

    412 ms

    Up

Alert quality

Paged at 3am for a network blip

After a few false pages, people mute the channel, and then the real outage lands in silence. A timeout or connection error first gets up to 4 attempts with backoff. On paid plans, if it still fails, WebPixie checks from the monitor's other selected locations. If the site answers there, no incident opens and the check stays out of your uptime figures. A monitor with one location has nowhere else to verify from, so the endpoints that page people should have at least two.

A wrong status code or a keyword mismatch is a real answer from the server, so it counts as down on the first check. For an endpoint that flaps, you can set the monitor to alert only after several failed checks in a row. Either way, you get one notification when the incident opens and one when it closes, not one for every failed check.

Check results

Last hour

  • billing.example.com

    Timeout · failed from 3 locations · 4 attempts

    CriticalIncident
  • app.example.com

    Timeout · reachable from New York · 4 attempts

    No alertUnverified

1 incident opened, 1 blip filtered out

SSL and certificates

The certificate on the subdomain nobody owned

An expired certificate on an API subdomain fails every client's TLS handshake at once. Browsers show a warning; integrations just fail. The usual cause is a subdomain that nobody's renewal script covered.

The Certificates Manager lists certificates seen for your domain and subdomains in Certificate Transparency logs, so a forgotten subdomain shows up with an Add Site action. Once it is a site, SSL monitoring validates the served certificate every day and warns ahead of expiry (15 days by default), raising the severity as the date gets closer.

It also compares the certificate on every IP the name resolves to. When one node behind a load balancer serves a different certificate, you see it as a finding to review.

Certificates · example.com

Daily check

  • staging.example.com

    Seen in CT logs, not monitored

    Add site
  • status.example.com

    SSL certificate expires

    6 days

    High
  • api.example.com

    Different certificate on 1 of 3 IPs

    Finding

Domains and DNS

A lapsed domain or a DNS change takes everything down

When a domain lapses, the website, the API and email go down together, and getting it back is slow.

Domain monitoring checks the registration every day and opens an incident as expiry gets close (30 days ahead by default) or when the domain is put on hold. Registrar status flags, such as the transfer lock, are shown next to it.

DNS monitoring checks more than 20 record types and DNSSEC every day and keeps each change with its old and new value, so when something breaks you can see what changed and when.

Domain and DNS · example.com

Daily check

  • example.com

    Domain expiration

    12 days

    High
  • MX · example.com

    mail.old-host.net → mx1.new-host.net

    Yesterday

    Changed
  • TXT · _dmarc.example.com

    p=none → p=quarantine

    3 days ago

    Changed

Incidents and alerts

Alerts that land in your on-call tool, not another inbox

On paid plans, each outage becomes one incident record with its type, severity, duration and how often it repeated, and your team can add notes. It closes on its own when checks pass again, which leaves you a timeline for the postmortem.

Alerts go out by email on every plan, in Slack from Starter and through webhooks from Pro, so incidents can go straight to the on-call tool you already use. WebPixie opens and closes the incident; your on-call tool decides who gets paged.

If you report uptime to customers, the percentage leaves out unverified failures and paused time, so it reflects real downtime.

Incidents

Last 7 days

  • billing.example.com

    Uptime · Timeout · 3 occurrences · 14 min

    CriticalOngoing
  • api.example.com

    SSL · SSL check · 38 min

    HighResolved
  • example.com

    Domain · Domain expiration · 2 days

    MediumResolved

Sent to Slack and your on-call webhook

API and MCP

Monitors that drift from what's deployed

A new service ships and nobody adds a monitor. With the GraphQL API on paid plans, the pipeline that deploys a service can create its monitor, pause it during the deploy and resume it afterwards. Paused time is left out of the uptime percentage, so a planned deploy doesn't count as downtime.

Keys can be read-only and limited to specific sites and uptime monitors, so a CI job only touches what it needs. The MCP server is open on every plan and lets Claude, ChatGPT, Codex or VS Code answer questions such as "what is down right now?" from your workspace.

GraphQL API

Deploy pipeline

New service: create its monitor

mutation {
  createUptimeLink(input: {
    target: "https://api.example.com/health"
    checkInterval: 300
    tags: ["api", "production"]
    http: { keywords: ["ok"] }
  }) {
    id
  }
}

Before the deploy: pause

mutation {
  pauseUptimeLink(ids: ["..."]) {
    affectedRecords
  }
}

One workspace

Uptime in one tool, cert expiry in a script nobody maintains

Every extra tool is another bill and another place to offboard someone, and none of them shows a subdomain's whole health.

In a WebPixie workspace, the uptime monitors, certificates, DNS records and domain data for every subdomain sit in one place. You can sort sites into folders per product, tag sites and monitors by service and environment, then filter to the one you are looking at.

Roles apply to the whole workspace: Owners handle billing and members, Admins manage monitors and incidents, and Users get read-only access. Every configuration change is recorded in an audit trail.

Sites

Filtered byproduction

API

  • api.example.comproductionScore91

App

  • app.example.comproductionScore86
  • auth.example.comproductionScore72

Marketing

  • example.comproductionScore88
  • docs.example.comproductionScore64

Setup in about a minute

No agent to install and no DNS changes.

  1. 1

    You add your services

    You add each domain as a site and each endpoint as an uptime monitor, by hand or through the API.

  2. 2

    You set the rules

    You pick the method, auth, accepted status codes, keywords, locations and interval for each monitor.

  3. 3

    You connect your channels

    Incidents go to email, Slack or a webhook your on-call tool receives.

Check intervals, monitor counts and alert channels differ by plan. You can compare them on the pricing page. Need a region beyond the defaults? On paid plans, support opens it within 24 hours.

Frequently Asked Questions

Common questions for DevOps teams using WebPixie.

On paid plans, WebPixie confirms the failure, opens one incident, sends one notification and closes the incident on its own when the endpoint recovers. For a timeout or connection error, the check is first retried in place, with up to 4 attempts in total, then verified from the monitor's other selected locations; a wrong status code or a keyword mismatch counts as down right away. The confirmed failure becomes an incident with its type, severity, start time and duration, and your team is notified by email and, depending on the plan, in Slack or through a webhook. While the incident is open, WebPixie re-checks the endpoint more often than its normal interval, so recovery is seen sooner and the outage length is measured accurately. When a check passes, the incident is marked resolved and a second notification goes out. Uptime incidents are always Critical, while SSL and domain expiry incidents start lower and move up in severity as the date gets closer. The incident record, with the notes your team adds, gives you the timeline for the postmortem. On the Free plan, you still get the email alert, but there is no incident list or history. You can see how checks work under uptime monitoring and the full lifecycle in the incidents docs.

WebPixie keeps network blips from opening incidents with two layers of protection: in-place retries for timeouts and connection errors, and verification from other locations on paid plans. First, when an uptime check hits a network-related failure such as a timeout or a connection error, WebPixie makes up to 4 check attempts with exponential backoff, so a brief blip does not open an incident. Second, on paid plans, if all attempts still fail, WebPixie runs verification checks from the monitor’s other selected locations before opening an incident. A monitor with a single location has no other location to verify from, so it is best to select at least two for the endpoints that page people. If the site is reachable from those locations, no incident opens and the failed check is recorded as unverified rather than confirmed down, and it is left out of your uptime percentage instead of counting as downtime. Failures that are not network-related, like an unexpected status code or a keyword mismatch, open an incident on the first attempt, because they reflect a real response from the server. For an endpoint that flaps, you can set the monitor to notify only after several failed checks in a row. You also control what counts as up: an accepted status-code list, keyword checks for content that must or must not appear, and a configurable request timeout all shape the verdict. When a failure is confirmed, it flows into incident management with a severity and a type, and you get one notification when the incident opens and one when it closes, not one per failed check. When checks recover, the incident closes automatically and is marked resolved. Per-plan check intervals are on the pricing page.

Yes. Each uptime monitor has a setting for how many failed checks in a row are needed before the alert goes out, and it is 1 by default. You can raise it for an endpoint that flaps, such as a service behind a slow upstream that returns an occasional 502, so a single bad answer stays out of your channels. The setting sits on top of WebPixie's own protection: timeouts and connection errors are already retried in place and, on paid plans, verified from the monitor's other selected locations before they count as a failure. A wrong status code or a keyword mismatch is not retried, so that is where the threshold helps most. A value of 1 suits endpoints where a single failure matters, like a login or payment route. The threshold holds back the notification, not the record: every check is still logged with its status code, location and response time in the uptime monitoring history. Once the alert goes out, the incident works as usual and closes on its own when checks pass again.

Yes. Alerts go out by email on every plan, in Slack from Starter and through outgoing webhooks from Pro, and a webhook is how you forward incidents to the on-call tool you already use. WebPixie sends a notification when an uptime, SSL or domain incident opens, when its severity changes significantly and when it closes, with email templates in English and Turkish. The same incident that opens in WebPixie is the one your channel receives, and it closes on its own when checks pass again. WebPixie detects and records the incident; your on-call tool decides who gets paged and how. Notifications are switched on or off per resource, so low-priority uptime monitors can stay quiet while production endpoints alert. By default, email goes to every member of the workspace. The channels included in each plan are on the pricing page.

WebPixie notices an outage on the next scheduled check, so the check interval you set for each monitor decides how quickly a failure is seen. The interval can be anything from 30 seconds to 24 hours, depending on your plan. Check options such as the method, headers, status codes, keywords and authentication are available on every plan. After a failing check, a timeout or connection error is retried in place and, on paid plans, verified from the monitor's other selected locations, which adds a short delay but keeps blips from paging you. A wrong status code or a keyword mismatch counts as down on the first check. The notification goes out as soon as the failure is confirmed, and the open incident is then re-checked more often so recovery is seen sooner. Endpoints that page people work best with a short interval, while low-risk pages can stay on a longer one. You can see how checks are confirmed under uptime monitoring and the interval for each plan on the pricing page.

Yes. Each uptime monitor can send HTTP Basic, Token or Digest authentication, custom request headers and a request body as form data, raw text or JSON, with the HTTP method you choose. You decide which status codes count as up, and a keyword check can require or forbid text in the response, so a health route that returns 200 with an error message still fails. A timeout of up to 15 seconds, a custom User-Agent and a follow-redirects switch round out the settings, and all of them are available on every plan; only the check interval depends on the plan. Every check records its location, response time, status code and response headers, and a failed check opens an incident on paid plans. You can compare intervals and monitor limits on the pricing page.

Everything WebPixie can watch for your services

You can turn on the checks each site and endpoint needs.

Related Solutions

You can put your critical endpoints under watch

Free plan, no credit card required.