Make your app fault tolerant by defining health checks and automatic restarts

Wouldn't it be great if your service could automatically restart when there is a deadlock or when it is not able to handle requests? Teams using LocalOps recover from service failures in seconds on their cloud or their customer cloud (BYOC) — without manual intervention.

Health checks:

Health checks when specified in ops.json probes your service containers periodically with the commands or endpoints you specify. If those checks fail continuously beyond a specified threshold, those specific containers will be automatically restarted.

This is highly useful if your code enters a deadlock or not executing at all for any reason. Configuring health checks and automatic restarts can detect such conditions and recover the application pretty quick. Quicker than your on-call SRE team to detect and fix.

Health checks can either be defined using shell commands like:

{
  "health": {
    "cmd": [
      "/bin/sh",
      "-c",
      "some_executable_or_script"
    ],
    "delay": 5, //seconds to wait, first time
    "interval": 5, //seconds between checks
    "failure_threshold": 3
  }
}

Or can have HTTP endpoint based checks like:

{
  "health": {
    "http": {
      "path": "/health", //optional; default: /;
      "port": 3000
    },
    "delay": 5,
    "interval": 5,
    "failure_threshold": 3
  }
}

Developer docs:

Refer to developer documentation for more details.

When you generate new helm charts using LocalOps for your enterprise self-hosting/on-prem customers, same probes are added there too for achieving fault tolerance!

Get Started Now

Sign up with LocalOps, connect your Github repo to deploy services on any cloud and define health checks in ops.json to make your applications more fault tolerant.

Schedule a personalized demo anytime to see how this can work for your services. We will get you running on the same day. Fun part — You don't even have to write Dockerfiles anymore to get started.

Auto-healing FTW.

Cheers.