Know your servers are healthy, not just your homepage.

ServerPal runs an agent on every host you own and checks every endpoint in front of them, in one place. When something breaks you already know which half it was.

  • 1 server and 20 checks free
  • No credit card
  • Checks from 30 seconds
acme-prod31
web-01app
cpu
mem
disk
web-02app
cpu
mem
disk
db-primarypostgres
cpu
mem
disk
cache-01redis
cpu
mem
disk

An agent on every server

One binary, installed with a single line. It registers against your project, reports within the minute and updates itself in place. Outbound only, so nothing to open on your firewall.

  • CPU utilisation, and a flag when the host is being throttled
  • Memory used against total, and every mounted filesystem tracked separately
  • Load average alerted per core, so one setting fits a 2-core and a 32-core box
  • Pending updates, reboot-required and failed systemd units, recorded but never alerted on

More on server monitoring →

db-primaryThreshold crossed
CPU64%
Memory78%
Disk /var88%
Load 1m7.41

Six kinds of check

Run them from one region or several, and set how many have to agree before an alert fires, so a single bad network path never pages anyone.

HTTP(S)

Status code, redirects and response-body keyword matching

Keyword

Fail when expected text is missing, even on a 200

TCP port

Confirm a service is accepting connections

Ping

Plain ICMP reachability

DNS

A record resolving to the value you expect

Heartbeat

Inbound pings from cron jobs and workers that stop silently

More on uptime monitoring →

Status pages your customers actually believe

Branded, on your own domain, and honest about the middle state: you set the share of components that has to be affected before the page stops claiming everything is fine.

  • Your accent colour and logo, set separately for light and dark
  • Slow counts as degraded when you want it to, not just hard outages
  • Opt-in email subscribers, double-confirmed, with per-plan limits
  • A read-only JSON endpoint per page on Pro and above

More on status pages →

status.acme.com
Degraded · 1 of 4 services affected
WebsiteOperational
APIDegraded
DashboardOperational
Background jobsOperational

Alerts that stay worth reading

Group your channels into reusable policies with delay, repeat and recovery rules, then attach a policy to any server or check.

EmailWebhookSlackDiscordTelegram
  • Repeated failures roll up into one incident, not a message per failed check
  • Public and internal updates on the same incident timeline
  • Maintenance windows publish themselves to your status page in advance
Incident #114
09:143 regions confirmed api.acme.com failing
09:16Posted to status.acme.com · 41 subscribers
09:52Recovered · duration 38m
EmailWebhookSlackDiscordTelegram

Simple, predictable pricing

Every plan includes every feature. What scales is how much you watch.

Compare every limit →

Common questions

Is the free plan really free?

Yes. One server, twenty checks, and a status page, with no card on file and no trial countdown. You upgrade when you outgrow it, not when a clock runs out.

What is the difference between a server and a check?

A server is a host running the ServerPal agent, which reports CPU, memory, disk, load and OS health from inside. A check runs from outside and tests an endpoint. They are counted separately, so installing the agent costs you no checks.

Do I have to open a firewall port for the agent?

No. The agent reports outbound on an interval and opens no listening port. There is no inbound rule to add and nothing for ServerPal to connect back to.

Can I put a status page on my own domain?

Yes, on any paid plan. Point a CNAME, verify it, and the page serves on your hostname with its certificate issued and renewed automatically.

Start monitoring in about a minute.

Twenty checks and one server, free and with no card. Upgrade only when you outgrow it.