HTTP Processing (Percent) gauge
The HTTP Processing (Percent) gauge monitors usage of available HTTP worker threads.
The gauge can raise alarms or generate a DEGRADED or UNAVAILABLE status that you can use to configure load balancing or auto-healing.
By default, this gauge raises an alarm at 70% usage, and it raises an alert at 90% usage. Also by default, the gauge does not mark the server as DEGRADED or UNAVAILABLE.
The following table explains the values and descriptions you set for this gauge.
| Value | Description |
|---|---|
|
This percentage value represents a warning condition. An alarm is raised, but the server continues to operate as normal. It defaults to 70%. |
|
This percentage value represents a severe condition. An alarm is raised, and the server enters a DEGRADED state. It is not set by default. To enable the DEGRADED state, you must set |
|
This percentage value represents a critical condition. An alarm is raised, an alert is generated, and the server is put into an UNAVAILABLE state. It defaults to 90%. To enable the UNAVAILABLE state, you must set |
|
The alarm level at which the server enters a DEGRADED state. By default, this gauge does not mark the server as DEGRADED. To enable the DEGRADED state, set to |
|
The alarm level at which the server enters an UNAVAILABLE state. By default, this gauge does not mark the server as UNAVAILABLE. To enable the UNAVAILABLE state, set to |
You can find the server’s availability state by using an option discussed in Server status.
The following example shows how to activate the gauge.
|
You might need to experiment to find values that work for your environment. |
Example
dsconfig set-gauge-prop --gauge-name "HTTP Processing (Percent)" --set major-value:85 --set server-degraded-severity-level:major --set server-unavailable-severity-level:critical