How is Service Health calculated when a device is in maintenance mode using Spectrum Service Manager

Products

• Network Observability

Issue/Introduction

The documents for Service Manager show Service Health is calculated with the following

 

Service Health Redundancy  
When all resources are Down then the service is Down.
When all resources are Degraded then the service is Degraded.
When all resources are Slightly Degraded then the service is Slightly Degraded.
When any 1 resource(s) are Down then the service is Slightly Degraded.

 

 

When we have 2 devices in a service that is being monitored under Service Health Redundancy, we have 1 device placed into maintenance and the other down.  We would expect to have only a minor alarm, but we receive a critical alarm.  Even though the device is in maintenance, it should still be counted as part of the Service Health Redundancy.

Environment

Spectrum with Service Manager : ANY

Cause

When a device is placed into Maintenance the code no longer calculates this device as being in part of the Service.

Resolution

Engineering has examined the code and has found that when a device is kept in maintenance, that device will be removed from the list of devices that contribute to service health. Hence when there are only two devices and one is in maintenance, service manager assumes it is monitoring only 1 device. As this is the only device it is monitoring,  it's service health changes as per that device condition. This is the reason customer is seeing service manger going to down state, when that condition of that only device going to critical state