Create an availability monitor

A monitor is a team-scoped active probe. Crystade checks your target from multiple geographic locations on an interval you choose. Any member can create, update, or delete a monitor.

Protocol fields, error codes, and timing metrics are in the monitor protocol reference. Limits are in plan limits.

Before you begin

• Select the team that should own the monitor. Monitors cannot move to another team.
• Use a public target. Private, loopback, and link-local destinations are rejected.

Choose interval and locations

Allowed intervals are 15 seconds, 30 seconds, 1 minute, 3 minutes, 5 minutes, 15 minutes, 30 minutes, 6 hours, 12 hours, and 24 hours. The plan minimum still applies: Free 5 minutes, Starter 1 minute, Standard 30 seconds.

Active monitoring runs from multiple locations at once. Supported regions are europe-north1, us-west1, asia-southeast1, and australia-southeast1. How many locations you can use is plan-gated (3 / 8 / 16).

The interval is a best-effort SLA. Execution can vary under high load. Schedule lag is kept under 5 seconds under normal load. Connection timeout is at most 5 minutes. A definition update takes effect within 5 minutes.

Create a monitor

Open Monitors to list uptime checks. Use New Monitor or Create a monitor.

HTTP

HTTP options include URL, request headers (max 20), request body (max 64 KB; plain text, JSON, or XML), redirect-follow limit (max 10), and connection timeout. HTTP/1.0, HTTP/1.1, and HTTP/2.0 are supported.

1. Open Monitors and choose Create a monitor (or New Monitor).

   

2. Set name, description, interval, and locations.

3. Choose HTTP and paste the URL.

4. Optionally set method, headers, body, redirects, and timeout.

5. Save, then link an assertion.

TCP

TCP options are host, port, and connection timeout. Results include time to DNS resolved (TTDR), time to first byte (TTFB), round-trip time (RTT), error code, and error message.

1. Create a monitor and choose TCP.

2. Set host and port (for example db.example.com and 5432).

3. Set timeout and locations, then save.

TLS

TLS certificate checks require Starter or Standard. Options are host, port, and connection timeout. TLS 1.2 and TLS 1.3 are supported. Results include the verified certificate chain, overall status, TTDR, TTFB, RTT, and TLS-specific errors.

1. Create a monitor and choose TLS.

2. Set host and port (usually 443).

3. After a check, inspect certificate subject, issuer, expiry, DNS names, and status.

Minecraft

Minecraft options are host, port, connection timeout, and protocol version. Results include version, MOTD, player count, max players, TTDR, TTFB, RTT, and handshake errors.

1. Create a monitor and choose Minecraft.

2. Set host, port (often 25565), and protocol version.

3. Save and confirm MOTD and player counts in the check log.

UDP monitoring is not available.

Read a check log

Each check log is immutable. It stores the monitor snapshot, probe location, per-protocol request and response snapshots, and timing breakdown.

All check logs are retained regardless of plan. How far back you can view them is plan-gated (30 / 90 / 365 days). Free does not capture HTTP response bodies; Starter captures up to 4 KB; Standard up to 16 KB. Upstream responses are capped at 32 MB.

Verify it worked

• The monitor appears in the team’s monitor list.
• Check logs arrive from the locations you selected.
• Timing fields (TTDR, TTFB, RTT) are populated on successful connects.
• Linked assertions produce healthy or failure runs. See Configure assertions.

Troubleshooting

Symptom	What to check
No checks yet	Wait up to 5 minutes after create or update. Confirm the monitor is not disabled or paused.
Interval rejected	Free cannot go faster than 5 minutes; Starter 1 minute; Standard 30 seconds.
TLS option missing	TLS checks require Starter or Standard.
Target rejected	Probes cannot target private, loopback, or link-local addresses. Use a publicly reachable host and port.

Related

Create your first monitor HTTP first-win walkthrough on your personal team.

Configure assertions Decide when a check is a failure and when an incident opens.

Monitor protocols Options, timing metrics, and error codes per protocol.

How monitoring relates to incidents The path from a probe to a public status update.