AWS::GameLift::ContainerGroupDefinition ContainerHealthCheck
Instructions on when and how to check the health of a container in a container fleet. When health check properties are set in a container definition, they override any Docker health checks in the container image. For more information on container health checks, see HealthCheck command in the Amazon Elastic Container Service API.
The following example instructions tell the container to wait 100 seconds after launch before counting failed health checks, then initiate the health check command every 60 seconds. After issuing the health check command, wait 10 seconds for it to succeed. If it fails, retry the command 3 times before considering the container to be unhealthy.
{"Command": [ "CMD-SHELL", "ps cax | grep "processmanager" || exit 1" ], "Interval":
300, "Timeout": 30, "Retries": 5, "StartPeriod": 100 }
Part of: ContainerDefinition$HealthCheck
Syntax
To declare this entity in your AWS CloudFormation template, use the following syntax:
JSON
{ "Command" :
[ String, ... ]
, "Interval" :Integer
, "Retries" :Integer
, "StartPeriod" :Integer
, "Timeout" :Integer
}
YAML
Command:
- String
Interval:Integer
Retries:Integer
StartPeriod:Integer
Timeout:Integer
Properties
Command
-
A string array that specifies the command that the container runs to determine if it's healthy.
Required: Yes
Type: Array of String
Minimum:
1 | 1
Maximum:
255 | 20
Update requires: Replacement
Interval
-
The time period (in seconds) between each health check.
Required: No
Type: Integer
Minimum:
60
Maximum:
300
Update requires: Replacement
Retries
-
The number of times to retry a failed health check before the container is considered unhealthy. The first run of the command does not count as a retry.
Required: No
Type: Integer
Minimum:
5
Maximum:
10
Update requires: Replacement
StartPeriod
-
The optional grace period (in seconds) to give a container time to bootstrap before the first failed health check counts toward the number of retries.
Required: No
Type: Integer
Minimum:
0
Maximum:
300
Update requires: Replacement
Timeout
-
The time period (in seconds) to wait for a health check to succeed before a failed health check is counted.
Required: No
Type: Integer
Minimum:
30
Maximum:
60
Update requires: Replacement