Metrics

Server metrics allow you to easily figure out when it’s time to scale. They can also be used, to identify issues with your code or bundle sizes, debug outages or detect memory leaks.

You can find server metrics by clicking on the server that you want to inspect and select the Metrics tab.

Core Metrics

At this point we provide three core metrics:

• CPU usage
• Memory usage
• Disk usage

CPU Usage

CPU usage tells you, how hard your server is working. You will see it going up, if your services do some heavy computations or the traffic on your website increases. If the CPU utilization reaches 100%, your server won’t be able to process any more requests, leading to infinite loading times on the client site and potentially breaking stuff. So always keep a close eye on the graph and if necessary scale your system before that happens.

Memory Usage

Each time your service executes a task it loads data into system memory in order to quickly work with it. The more data your applications process the more memory they require. Same as with CPU, this capacity is limited so when you reach 100% stuff might break.

Disk Usage

The disk metric shows how much space is left on your server. Each image, container and the data that they store eat up disk space. If the space is full, deploys will stop working and your apps won’t be able to save any more data.

Clear Cache

Disk usage includes cached layers from Docker builds that happened on the server. The cache will be freed automatically if the disk space reaches more than 60% of your servers storage capacity.

You can manually clear the build cache by clicking the “Clear Cache” button.

Per Service Metrics

By default, the metrics graphs show you the total system usage.

You can also break down resource utilization of each service individually, by selecting the service that you want to inspect, from the filter view next to the metrics graphs.

Keep in mind, that the sum from all services does not add up to the total system usage, since total system usage also includes other processes that are running on the server, like the web server, the Docker daemon and other OS packages.

Timeframe

You can choose between five different timeframes, allowing you to inspect metrics of the last

• 10 Minutes,
• 1 Hour,
• 12 Hours,
• 24 Hours,
• 3 Days.

Limits

On development servers, your server metrics are limited. In order to fully experience metrics capabilities, upgrade your server to a production instance.

You can learn more about the two different server types development and production in chapter: Create a Server.