Health Checks can be set up per Ignite deployment to make sure that:
- Traffic is only directed to containers that pass the check, and
- New builds smoothly roll out without downtime
This is done by querying the containers periodically to make sure they're alive and healthy. You can choose how to respond to a health check. For example, you may only want to respond successfully to a health check once your program has connected to its database.
|Path||the HTTP path that should be queried to check if your service is healthy|
|Port||the port that your service is running on|
|Initial Delay||5||How long Ignite should wait before sending the first health check|
|Interval||60||How long Ignite should wait (in seconds) between sending health checks|
|Timeout||10000||The maximum time, in milliseconds, that should be tolerated for receiving a response to the health check. If exceeded, the container will be determined as unhealthy|
|Maximum Failures||3||How many times the health check needs to fail for a container to be determined unhealthy|
Currently, Ignite only supports HTTP(s)-based health checks. When setting up a health check, you can choose the HTTP path to be queried to make sure the container is alive.
To respond successfully to a health check, you must respond on the path specified when creating the health check with a
HTTP 2xx response code, such as
Navigate to an Ignite deployment
Select Health Check from the sidebar
Click Create Check in the top right