When running complex online systems, the last thing you want to experience is a server crash. Thus, it’s important for professionals to conduct server health checks in order to keep them up and running. Some health checks can be difficult to incorporate. Here are the strengths and differences between various health check systems.

1. Basic Checks

These checks are commonly performed by a load balancing software or external monitoring agent in order to test the basic connectivity to a service. They are often included with the service and don’t require that an application author implement anything, nor does its user need to be aware of how the application itself works. Basic checks include:

• Confirming that a server is accepting new TCP connections and listening on its expected port.
• Performing basic HTTP requests and ensuring that the server is responding with a 200 status code.
• Reviewing statuses and whether the system is operating properly by testing features such as network availability.

2. Local Health Checks

Going beyond basic checks, local health checks test resources that aren’t shared with the server’s peers. The complexity of the system keeps it from failing on servers in the pool that function simultaneously. These tests check for:

• The inability to write to or read from disks.
• Whether critical processes are crashing or breaking.
• Whether there are missing support processes.

3. Dependency Health Checks

A far more thorough inspection of an application’s ability to interact with adjacent systems, dependency health checks are ideal for checking up on local servers. If there are problems with the dependency itself, these health checks can sometimes result in false positives. Dependency health checks review:

• Bad configurations or stale metadata such as broken update mechanisms that cause servers to become out of sync with their peers and behave in unpredictable and untested manners.
• An inability to communicate with peer servers or dependencies caused by strange network behaviors or software issues such as deadlocks or bugs.
• The presence of unusual software bugs such as memory leaks or state corruption bugs.

4. Anomaly Detection

Anomaly detection takes into consideration every server in a in a pool, aggregating this data in order to continuously compare error rates and latency data. It can discover divergences that a server cannot detect itself such as:

• Clock skews, which occur when servers experience a high load and the server’s clock skews abruptly. These can cause requests to services to fail if client clocks aren’t synced with the server.
• Whether a server has become disconnected from a network or turned off because it’s running outdated code.
• Unanticipated failures that are identified as a client error. Servers might be returning false responses to their callers.

Load Balancing with Server Health Checks

Enterprise data can dictate the success of your business, so you’ll want to ensure that your applications are responding well and always available with a load balancing software. A lack of proper balancing can put your company at risk of stressed servers and a lack of workplace efficiency. Load balancing can maximize your performance, save you on costs, and allow you to monitor your servers 24/7.

Advanced Server Load Balancing Systems

At Resonate, we help you evenly distribute service requests with advanced and efficient technology. Work with us now to get better use out of your resources and manage them in ways that are dynamic and scalable.