Watching 6 servers · nothing needs you1 of 6 needs you · db-01 · diskTue 22:40Wed 03:02Wed 03:14Wed 03:14
Know your servers are fine. Hear about it when they’re not.
Server monitoring for teams running dedicated or cloud infrastructure. One command to install, nothing to host, and alerts only when something needs a human. The full picture is there when you want it.
The fleet at a glance: who is online, what needs attention, and the open alerts, sorted so problems come first.
app.monitorable.io/servers/db-01
db-011H3H24H7D30D
CPU
47%
8c/16t · load 1.9 · 51°C
Memory
72%
11.5 GB / 16 GB
Disk
62%
149 GB / 240 GB
Swap
0%
0 B / 4 GB
Hostname
db-01
OS
Ubuntu 24.04.4 LTS
Kernel
6.8.0-139-generic
IP
198.51.100.7
Location
Finland
CPU model
AMD EPYC 7502P · amd64
Agent
v1.4.0
Uptime
112d 2h
Last check
Just now
Alerts on this server
No open alerts
Mute alerts
Pauses evaluation for this server
Template
Account default
DiskUsageI/OIOPSBusy-1h … now
/ (root)
CPUUsageLoadTemperature-1h … now
TotalUserSystem
app.monitorable.io/servers/db-01
db-011H3H24H7D30D
CPU
47%
8c/16t · load 1.9 · 51°C
Memory
72%
11.5 GB / 16 GB
Disk
91%
218 GB / 240 GB
Swap
0%
0 B / 4 GB
Hostname
db-01
OS
Ubuntu 24.04.4 LTS
Kernel
6.8.0-139-generic
IP
198.51.100.7
Location
Finland
CPU model
AMD EPYC 7502P · amd64
Agent
v1.4.0
Uptime
112d 2h
Last check
Just now
Alerts on this server1
Disk usage · /
firing · 6m · rule disk_usage
Ack
Mute alerts
Pauses evaluation for this server
Template
Account default
DiskUsageI/OIOPSBusy-1h … now
/ (root)
CPUUsageLoadTemperature-1h … now
TotalUserSystem
One server on one page: vitals, system details, its alerts, and every chart, from the last hour back to a year.
app.monitorable.io/alerts
AlertsUpdated 2s agoIncidentsTemplatesChannels
0 alerts firing
0 firing0 pending0 acked
Alerts · 24h
1
1 auto-resolved
Mean time to ack
—
last 7 days · 0 acked
Mean time to resolve
6h 2m
last 7 days · 1 recovered
All alertsAll 1Firing 0Acked 0Resolved 1 Filter by alert, server, rule…
Alert
Server
Started
Duration
Status
Actions
Disk usage · /
rule disk_usage
03:14
6h 2m
RESOLVED
resolved · 09:16
app.monitorable.io/alerts
AlertsUpdated 2s agoIncidentsTemplatesChannels
1 alert firing
1 firing0 pending0 acked
Alerts · 24h
1
0 auto-resolved
Mean time to ack
—
last 7 days · 0 acked
Mean time to resolve
—
last 7 days · 0 recovered
All alertsAll 1Firing 1Acked 0Resolved 0 Filter by alert, server, rule…
Alert
Server
Started
Duration
Status
Actions
Disk usage · /
rule disk_usage
03:14
6m
FIRING
Ack
Every incident, firing or resolved, with how long it took to be seen and to be fixed.
What one agent covers
Linux only: x86-64 and ARM64, on bare metal, VMs, and LXC. The full list of what leaves your server is in the docs.
The host
CPU, load, memory, swap, disk usage and I/O, network, temperatures. Every 60 seconds.
Docker containers
State, health checks, restarts, CPU and memory per container. Detected automatically.
systemd services
Failed units, restart loops, resource use per service.
Physical disks
SMART health, wear, new errors and reallocated sectors, on bare metal.
History
Per-minute detail for weeks, hourly for up to a year, depending on plan. Charts keep the real peaks when zoomed out.
Alerting
19 rule types, 12 on by default. Templates per server. Email, Slack, Telegram, Discord, PagerDuty, ntfy, and webhooks.
04Why people use itFour complaints, four answers
Built for the parts of monitoring that usually go wrong
“01
I don’t want to keep a dashboard open all day.
You don’t have to. Alerts reach you where you already are. The dashboard is there for when you want the full picture, not as a place you have to sit.
“02
Noisy alerts get muted, and then nobody sees the real one.
Thresholds are averaged over a window and need two checks in a row, so a single spike sends nothing. Reminders are capped per rule, a flapping condition stays one incident, and acknowledging stops the noise.
“03
A server that goes silent is the one I find out about last.
A silent agent is an alert of its own, on by default, five minutes after the last report. A server that stopped reporting and a server that is down are shown as different things.
“04
When something fires, I need context, not a code.
Every alert names the server, the disk, unit, or container, the number, and the line it crossed, and links to that server’s page with the charts around that moment.
05Nothing hiddenApache-2.0 · OTLP · EU
Check for yourself, rather than take our word
There are no customer logos or testimonials on this page yet, and we won’t make any up. What we can offer is everything you need to verify the product on your own terms.
An open-source agent on an open standard
The agent is a single static binary built from public source under the Apache-2.0 license. It is a real OpenTelemetry Collector distribution, so your metrics leave the server as standard OTLP: nothing proprietary runs on your machines, and you never need to know any of that to use it.
It reads metrics and pushes them out over HTTPS. It never opens your logs or your files. You can confirm that in the source.
Run in the EU, by a person you can email
The platform runs on dedicated servers we operate in France. Sign-in is GitHub, Google, or an email link. Tenancy is enforced on every query.
Every monitoring setup I ran ended the same way: a stack I babysat more than the servers it watched, and an alert channel everyone had muted. Monitorable is the tool I wanted instead. If it falls short, write to me. I answer.
Pick the tier that fits your fleet and pay one flat price for it. Nothing is metered: not metrics, not containers, not alerts. Three servers are free forever; every paid tier includes a 14-day free trial, no card required.
With a tier full, that is $0.90 a server on Pro, $0.73 a server on Scale and $0.53 a server on Enterprise. Prices in USD.
07FAQThe questions that block a decision
Before you decide
Do I need to know anything about OpenTelemetry?
No. The agent is an OpenTelemetry Collector build, but you never configure it: install, done. If you already use OTLP tooling, your metrics are in the format you expect.
How heavy is the agent?
Light, and measured: about 47 MB on disk, 12–16 MB of memory, and under 0.1% of one core on the server the docs are verified against. A single static binary, no runtime.
What happens if a server stops reporting?
It is marked Stale, deliberately distinct from Offline, and the built-in Collector offline rule notifies you after five minutes without a report. A gap in monitoring is never silent.
Does Monitorable connect to my servers?
No. The agent pushes metrics out over HTTPS, and nothing connects in. There are no external probes, no inbound ports, and no remote commands.
Which platforms are supported?
Linux, on x86-64 and ARM64: bare metal, VMs, and LXC. That focus is why setup is one command. Other platforms are planned, without a date.
How is this different from an uptime checker or a big observability platform?
Uptime checkers tell you the site is down. The big platforms tell you everything, priced in dimensions you need a spreadsheet to predict. Monitorable is the middle that was missing: deep server health, one flat monthly price, five-minute setup.
Can I get my data out?
Metrics leave your server as standard OTLP, so nothing about the format is proprietary. There is no self-service export yet; write to hello@monitorable.io and we send you a copy.
How do promo codes work?
Enter the code in the checkout dialog when you subscribe; the discounted amount is shown before you pay and appears on the checkout page and your receipt. Codes apply to the plans and period they were issued for.
08Get startedInstall · stop the agent · watch the alert arrive
Put it on one server tonight
Install on a box you care about. Stop the agent if you like, and watch the offline alert arrive five minutes later. If it isn’t the easiest monitoring you have set up, removing it is one command too.