Active IQ Unified Manager Discussions

Source of Alerts in Balance

MAHESH1111111
3,009 Views

Does anyone know if we can configure custom alerts in balance For Ex: If a particular volume crosses 100 IOPS - trigger an alert or if the Response time is more than 15ms - trigger an alert.

I could see the default alerts page in balance - but wanted to know if we can create our custom alerts based on our needs. Any Suggestions/ thoughts would be of great help. Thanks

Regards

Mahesh.P

1 REPLY 1

dbourque
3,009 Views

Hello Manesh,

Although OnCommand Balance captures and utilizes much of the requested data for identifying performance bottlenecks and contention in infrastructure,  there is no mechanism in the Balance product today to manually configure for a specific iop or response time threshold. Balance is a trend analysis infrastructure performance monitoring tool (not real time) and leverages that data (hourly averages) along with its proprietary analytics and modeling as its approach to infrastructure performance management. Products such as Insight Perform will likely allow you to trigger on such values. Balance focuses on latency (Q-depth) through out the various layers of the Data i/o path and calculates the concept of headroom and not solely alerts on pure iops reads/writes or an arbitrary value in response time.  For example,  a way to think of it is; if the device is driving 1000 Iops is it good or bad?  Is the device is able to to get the Iops needed without impacting a different workload/server at that period of time and the response time/service latency is not going up.

All that said, under the config tabs throughout the product,  you will find options to manually set the thresholds for CPU % warnings, Memory pages per second, Controller cpu utilization %, Capacity thresholds for volumes, # of weeks to forecast and under the Applications tab, you can set the perceived Disk Utilization %.

Lastly, many of the scorecard reports can be leveraged as well to provide you and operational view of max, min and average values for Iops, capacity, q-depth, response time etc.

I hope the above mentioned is helpful.

With Best regards,

Don Bourque

Insight SE

Netapp

Public