We Have Moved
The IBM Application Gateway has a new home - docs.verify.ibm.com/gateway
The documentation on this site will no longer be maintained after v21.02, please update your bookmarks.
Health
Description
Specifies the configuration which is used to establish the health of the resource server.
Properties
The following table(s) describe the configuration properties for this component:
| Name | Type | Constraints | Description |
|---|---|---|---|
| ping | PING Object |
PING Object
Specifies the configuration that allows the reverse proxy to periodically ping each resource server to determine whether it is running.
| Name | Type | Constraints | Description |
|---|---|---|---|
| url | string | The URL to which the ping request will be sent. |
|
| policy | POLICY Object | ||
| method | string | The HTTP method which will be used in the ping request. |
|
| rule | array[string] | Regular (client initiated) requests can also be observed to determine the resource server health. This entry is an ordered list of rules based on the response status codes. Status codes prefixed with a '+' are considered healthy, and codes prefixed with '-' unhealthy. The wildcard characters '*' and '?' can be used. |
POLICY Object
The policy associated with the ping requests.
| Name | Type | Constraints | Description |
|---|---|---|---|
| threshold | number | Minimum:1 Maximum:unlimited Default:1 |
Specifies how many consecutive times the ping must fail before the resource server is considered unhealthy. |
| timeout | number | Minimum:0 Maximum:unlimited |
Specifies how long the reverse proxy should wait for responses to ping requests. A value of 0 will result in an indefinite wait. If not specified the timeout will default to the http or https timeout. |
| frequency | number | Minimum:0 Maximum:unlimited Default:300 |
The frequency (in seconds) which ping requests will be sent. If set to 0 the ping will be disabled. |
| recovery | RECOVERY Object | ||
| rule | array[string] | Specifies how to interpret responses to ping requests. This entry is an ordered list of rules based on the response status codes. Status codes prefixed with a '+' are considered healthy, and codes prefixed with '-' unhealthy. The wildcard characters '*' and '?' can be used. |
RECOVERY Object
When the resource server is unhealthy, these entries specify how often they should be pinged and how many consecutive successful pings should be observed before the resource server is considered healthy again.
| Name | Type | Constraints | Description |
|---|---|---|---|
| threshold | number | Minimum:1 Maximum:unlimited Default:1 |
Specifies how many consecutive times the ping must fail before the resource server is considered unhealthy. |
| frequency | number | Minimum:1 Maximum:unlimited |
The frequency (in seconds) which ping requests will be sent. If this value is not set it will default to the value of the ping frequency. |
Example
resource_servers:
- path: "/example"
...
health:
ping:
method: HEAD
url: /
policy:
frequency: 600
threshold: 1
recovery:
frequency: 300
threshold: 1
timeout: 10
rule:
- -50?
- +*
rule:
- -50?
- +*