Opyntel
Coolify

Coolify monitoring with the CPU and memory alerts Sentinel does not send

Opyntel’s Coolify monitoring reads deploys, restarts, resources and alerts for every application and server on your Coolify instance. It only reads, and nothing changes in your Coolify install.

See how it works
Built in vs Opyntel

Coolify on its own, and Coolify with Opyntel

Feature by feature, the biggest gaps first: what Coolify gives you, and what you get in one dashboard with Opyntel.

All your services in one dashboard

On Coolify

One Coolify instance manages resources on one or more servers. Its CPU and memory charts sit under each server (Servers > [server] > Charts) and on each running application or standalone database, and the docs say resource metrics are not available for Docker Compose applications or one-click services. The multi-server docs describe no combined monitoring view and list monitoring for the application and every server as something you set up yourself.

With Opyntel

Opyntel puts the full list of your services in one table: every application, database and server from every Coolify instance, down and degraded first, grouped by project. You see it all at a glance, Coolify plus the other systems you run, with no clicking from server to server.

Latency percentiles and error rate

On Coolify

Traffic analytics shows requests, bandwidth, visitors, p95 latency and a status-code breakdown per server, application or service. It is off by default and enabled per server, which restarts the proxy, and the page lists no error-rate figure or other percentiles.

With Opyntel

Opyntel shows requests, status classes, p50, p95 and p99 latency and the 5xx error rate for every public service, side by side in the same table. Slow or failing services stand out next to the healthy ones without opening each one.

Alerts when it goes down, and when it recovers

On Coolify

Coolify notifies on deploys, resource status changes, restart limits, backups, scheduled tasks, disk usage warnings and a server becoming unreachable or reachable again, by email, Discord, Telegram, Slack, Pushover or webhook. The monitoring overview lists CPU or memory threshold alerts from the built-in charts as needing an external system.

With Opyntel

Opyntel is that external system: it alerts when a service goes down, when CPU, memory, error rate or p95 latency passes your threshold, and when a disk reaches 85% or 95% or is on track to fill, with custom rules per service, project or environment on Pro. Every alert notifies when it fires and again when it resolves, by email, Slack or SMS depending on your plan.

History

On Coolify

Sentinel keeps CPU and memory history for 7 days by default, stored locally on each server. Traffic analytics keeps hourly and daily aggregates for a number of days you set.

With Opyntel

Opyntel keeps the same CPU and memory history, and the request metrics beside it, for 30 days on the free plan, a year on Pro and unlimited on Enterprise. It sits outside your servers, so it stays when a server is rebuilt or removed.

Deploys on one timeline

On Coolify

Coolify can send a notification for each deployment success or failure to the channels you pick. Each channel has its own event selection.

With Opyntel

Opyntel puts every deploy with its commit on one activity timeline across all your Coolify servers and instances and your other platforms. When a service turns red you can see what shipped just before.

The total cost of your app

On Coolify

Coolify is not a hosting provider for workload servers: the docs list choosing and paying for the connected servers as your responsibility. The bill sits with each server’s host.

With Opyntel

Opyntel shows spend per server and project on Pro, with list-price estimates for VPS servers, next to the spend of the other systems you run. You see the total cost of your app at a glance, with monthly budget alerts and forecasts on Enterprise.

The Coolify dashboard

The Coolify dashboard: what it shows and what it misses

Coolify’s own dashboard is where you deploy, read logs and set up the proxy. It is built around one instance and the servers it manages.

What the Coolify dashboard shows

  • Deployments for each application, with their status and build logs
  • Application and container logs
  • Sentinel CPU and memory graphs on each server’s Charts page, once metrics are on
  • Proxy configuration (Traefik or Caddy) for each server

What it does not show or alert on

  • Alerts when CPU or memory crosses a threshold
  • Several Coolify instances in one dashboard
  • A signal when Coolify itself is down: the dashboard is the thing you cannot reach
  • Up to a year of history: metric retention is set per server, 7 by default, on that server’s own disk
Coolify vs Opyntel

What the console gives you, and what Opyntel adds

Coolify sends deploy, restart, server and disk notifications to a wide choice of channels. Opyntel adds resource alerts, history and one view across servers.

CapabilityThe Coolify consoleWhat Opyntel adds
Deploy and restart notificationsThe Coolify console

The docs list deployment success and failure, resource status changes (a container stopping unexpectedly) and restart-limit events, sent to email, Discord, Telegram, Slack, Pushover or a webhook. In practice, users report no notification when a container is killed or crashes.

What Opyntel addsDeploys and restarts on one activity timeline, and an alert when a service goes down. Keep Coolify's deploy notifications alongside it.
CPU and memory thresholdsThe Coolify console

Historical CPU and memory charts when Sentinel metrics are on. Threshold alerts for CPU or memory are not built in and need an external system.

What Opyntel addsCPU and memory alert rules on the per-container CPU and memory Sentinel collects, scoped to a project, environment or service.
Disk usageThe Coolify console

A disk usage notification when server disk usage passes the threshold set for that server, repeated every 24 hours by default. Sentinel reports root filesystem usage.

What Opyntel addsDisk rules on the host disk, including a forecast of when it fills, and the usage history behind them.
Server reachabilityThe Coolify console

A Server Unreachable notification after two consecutive failed checks, and Server Reachable when it recovers.

What Opyntel addsServer state in the same table as your applications, and an alert when Opyntel cannot collect from your Coolify instance.
Multi-server viewThe Coolify console

Sentinel and its metrics are set up server by server, on each server’s Charts page.

What Opyntel addsServers, applications and databases from your Coolify instance in one table, filtered by project or environment, beside anything else you connect.
History lengthThe Coolify console

History retention is a setting on each server’s Charts page (the default is 7) and is stored on that server’s own disk. A 30-day chart shows only the history still available.

What Opyntel addsUp to a year of hourly metric history, kept outside your server. Opyntel stores metrics, so logs stay in Coolify.
Read-only accessThe Coolify console

The read permission lists and inspects resources and keeps sensitive values redacted. Write and deploy are separate permissions.

What Opyntel addsOpyntel uses the token only to read, enforced in code, and stores it encrypted.
Alerts by plan

What Coolify alerts on, self-hosted or Cloud

Coolify has one free self-hosted edition and Coolify Cloud, which runs the dashboard for you while your servers stay yours. Both send the same Coolify notifications.

AlertSelf-hostedCoolify CloudOpyntel
Deploy failedSelf-hostedYes, to email, Slack, Discord, Telegram, Pushover or a webhookCoolify CloudYes, with email alerts set up for youOpyntelFailed deploys on the activity timeline and the daily grid, beside the metrics around them.
Container stopped or crashedSelf-hostedListed in the docs, but users report none on a kill or crashCoolify CloudSame Coolify code, same gapOpyntelA service-down alert when a service stops or crashes, with restarts on the timeline.
CPUSelf-hostedNo threshold alert; charts only, with Sentinel metrics onCoolify CloudNo threshold alertOpyntelA default high-CPU rule on the per-container CPU Sentinel collects, scoped to a project, environment or service.
MemorySelf-hostedNo threshold alert; charts only, with Sentinel metrics onCoolify CloudNo threshold alertOpyntelA default high-memory rule on the per-container memory Sentinel collects.
DiskSelf-hostedYes, when server disk usage passes its thresholdCoolify CloudYes, same notificationOpyntelDisk filling, almost full and a forecast of when it fills.
SpendSelf-hostedn/a, Coolify is freeCoolify Cloudn/a, a flat price per serverOpyntelMonthly cost alert rules on the monthly cost you enter for Coolify and on host spend where a host connector reports it.
Certificate expirySelf-hostedNot among the documented notification eventsCoolify CloudNot among the documented notification eventsOpyntelAn outside-in check that alerts before a TLS certificate expires, at the number of days you set.
Costs

Coolify is free. The server underneath is not

Self-hosted Coolify costs nothing and Coolify Cloud charges a flat price per server. The bill that moves is the one from your server provider.

What you pay for
The VPS or dedicated server Coolify and your apps run on, plus a flat Coolify Cloud fee per server if you use Cloud.
Where surprises come from
Disk that keeps growing with images, volumes and logs, backups kept on the host or in storage, and bandwidth billed by the host.
What Opyntel shows
Server CPU, memory and disk next to the apps using them, and host spend where a host connector reports it, such as an estimate from Hetzner Cloud list prices.
Budgets
A monthly cost rule that alerts when spend across your connected sources passes the budget you set.
Setup

How to set up Coolify monitoring with Opyntel

Coolify monitoring in four steps: from Coolify to one dashboard with everything else you run.

  1. Create a Coolify API token with the read permission
    In Coolify, open Keys & Tokens, then API Tokens, and create a token with only the read permission, never write or deploy. On a self-hosted instance, an administrator enables API access first.
    Create a Coolify API token
  2. Sign in to Opyntel
    Open your Opyntel dashboard and add a source. Accounts start from the waitlist, so join it to get yours.
  3. Paste the token in your dashboard
    Add Coolify as a source and paste the token with your instance URL. For CPU and memory, add the Sentinel token from each server’s Sentinel page as well.
  4. See Coolify in one dashboard
    Coolify monitoring lands next to the other platforms you run: every service and server sits in one dashboard, down and degraded first, with default alert rules on from the start.

Don’t see your platform? From your Opyntel account you can request an integration with more platforms. Until then, and tell us which one you run on.

Dashboard

All your Coolify resources in one dashboard

All servers and projects in one table.

Opyntel

Applications, databases and servers from your Coolify instance land in one table, down and degraded first, with CPU, RAM and the last deploy. Open any one for its status history, events and usage charts.

Services

Down and degraded first

Example data
Services
6
Down or degraded
2
Servers
1
Last sync
40 s ago
ServiceStatusProjectSourceCPURAMLast deploy
workerdownshopCoolify——2 h ago
apidegradedshopCoolify0.84 vCPU640 MB18 min ago
webdeployingshopCoolify0.05 vCPU180 MBjust now
postgresokshopCoolify0.22 vCPU1.1 GB6 d ago
docsokblogCoolify0.02 vCPU96 MB4 d ago
server-2okserversCoolify—5.2 GB—
FAQ

Coolify questions, answered

Sentinel, tokens, restarts and network access.

For CPU and memory, yes. Opyntel reads those from Sentinel, and Coolify leaves metrics off until you enable them on the server’s Charts page. Without Sentinel metrics you still get states, deploys and restarts, but no CPU or memory alerts. Sentinel is not available on build servers or Docker Swarm servers, and Coolify has no metrics for Docker Compose applications or one-click services.

Yes. Opyntel talks to the Coolify API, so it does not matter which provider the server runs on. A Coolify instance on a Hetzner server works like any other, and its CPU, memory and disk come from Sentinel.

Yes. Hostinger sells VPS plans with Coolify preinstalled (Hostinger’s Coolify VPS guide). Connect it with a Coolify API token like any other instance, and run the Opyntel agent on the VPS for host CPU, memory and disk, and container restart counts.

No. Opyntel only reads, which is enforced in code. Use a token with the read permission only, without write or deploy, so Coolify itself refuses anything else.

Only if its API is reachable from the internet. Opyntel calls your Coolify API from outside, so a Coolify instance that accepts no inbound connections cannot be connected with a token.

No. Coolify’s notifications for deploys, restarts, server reachability and disk usage work well, and you can keep them. Opyntel adds CPU and memory alerts, a longer history and one view across servers, by email, Slack or SMS depending on your plan.

On connect, Opyntel loads the last 7 days of Sentinel metrics, then keeps hourly history for up to a year. Opyntel stores metrics and not logs, so your logs stay in Coolify.

The kernel killed it for memory, and with no limit set, the whole server can run short. How to confirm and fix exit code 137 on Coolify.

Its container exits and is started again until Coolify’s restart limit stops it. The logs and exit code show the cause. Why a Docker container keeps restarting.

Traefik asks Let’s Encrypt for a certificate once the domain resolves to the server and port 80 is reachable; until then it serves its own default certificate. Opyntel alerts before a certificate expires. Fix an expired or missing SSL certificate.

The container reached its open files limit. Raise it with ulimits in the Compose file or custom Docker options, then check for a leak. Opyntel’s Coolify monitoring alerts when open files pass your threshold. Fix too many open files in a container.

Each scheduled task lists its executions and output on the resource’s page. Opyntel’s Coolify monitoring alerts when one fails or does not run. Why a cron job is not running.

The alerts your platform doesn’t send.

One dashboard for everything you run, with the alerts your platforms leave out. Read-only, nothing to install in your apps.

See how it works