Researches on NetApp OnCommand Balance Part 2 - Some Use Cases

Carrying on from this post of two months ago -
- here we expand the section ‘Use Cases’

Some Use Cases of OnCommand Balance:

Use Case 1: Find All Servers That Share Groups or Aggregates

Balance > Storage Arrays tab > Select the Storage Array
Performance Summary tab > Select the Disk Group or Aggregate
At the bottom of the disk group summary page, notice the breakdown of all the workloads that share the disk group or aggregate.

Data Topology tab >
For a more visual understanding of the workloads that share the disk group.

Use Case 2: Identify Application Workloads That Are Limited by CPU, Memory, or I/O Availability

Balance > Applications tab >
This page list all the workloads and status indicators for their CPU, Memory, and Storage (I/O) resources. View the resource status icons.

Balance > Applications tab > Select the Application
To get a more detailed look at a specific application. The infrastructure response time graph shows the end-to-end response time for a workload, including its CPU and storage.

Data Topology tab >
This provides a visualization of the application workloads path to storage. If the workload is a database application (Microsoft SQL or Oracle), click the schema check box to view the full data path for each database element.

Balance > Reports > Scorecard Reports > Application Scorecard
The Application Scorecard provides response time, CPU, memory, and storage utilization information for application workloads. Generally, overall response times should not exceed 100ms!

Use Case 3: Identify Bully and Victim Servers

Open and read the Balance predictor email.
For further analysis, click the View full analysis link and view the workload breakdowns.

Use Case 4: Reclaim Storage and Identify Storage That Is Nearing Capacity

Balance > Reports > Standard Reports > Server Volume Capacity Forecast
Balance > Reports > Scorecard Reports
Balance > Reports > Reports Scheduling

To see volumetric capacity at the array level, use the Array Utilization Report.
To see capacity at the disk group or aggregate level, use the Storage Scorecard.
To see current capacity from the perspective of each server volume, use the Server Storage Utilization Report.
To see capacity from the perspective of application workloads, use the Application Storage Trend Report.
The Server Volume Capacity Utilization Forecast Report provides the calculated number of weeks before server volumes hit 80%, 90%, and 100% capacity.

Capacity information is available throughout the UI as you view an array with its disk groups, or drill down for data on a specific disk group, or analyze the server volume for detailed capacity information. Balance array performance reports show volumetric capacity information in addition to high-level array information, and detailed statistics and analysis of the arrays disk groups, ports, and controllers.

Use Case 5: Identify Servers That Are Causing Resource Contention

At the host level:
Balance > Servers > Virtual Hosts > select a virtual host
The Summary page provides quick access to contention information for the shared CPU and memory.
Tick the CPU ‘Show VM breakout’ box - to access the VM breakout for CPU usage. The display includes hyperlinks to access the summary page for each VM.
Tick the Memory ‘Show VM breakout’ box - to access the VM breakout for Memory usage. The display includes hyperlinks to access the summary page for each VM.

At the disk group (aggregate) level:
Balance > Storage > select an array
Disk Groups tab > select a disk group
The workload breakdown at the bottom of the Summary page, displays the performance characteristics for each contending workload (sort by: Most IO, Worst RT, Highest disk utilization.)
See the Contention tab - this display includes the disk utilization percentage, response time and throughput.
See the IO by Workload tab.
Also, see the Scorecards and Reports, for example - the Virtual Machine Scorecard report, the Application Scorecard report

Note: Keep in mind that reports can be scheduled!

Use Case 6: Identify Storage Hotspots

Balance > Dashboard
The Balance Dashboard proactively identifies storage hotspots. Click the link for Arrays and check the Status column for anything that is in red. The dashboard also lists the most recent abnormal storage events.

Balance > Storage > select an array
Data Topology tab >
Red symbols indicate hotspots. For a specific element you can right-click and choose ‘Re-orient Topology’. Or, right-click and choose ‘Open Summary Page’.

Balance > Reports > Scorecard Reports > Storage Scorecard
Use to identify disk groups with excessive disk utilization, response times, and IOPs; in addition to capacity information.

Use Case 7: Identify Servers That Have Misaligned LUN Partitions

Balance > Servers
The top of the page indicates the total number of servers that have misaligned partitions. Click ‘View the report’ and use the ‘Servers With Misaligned Partitions’ report to focus your efforts on those servers that are misaligned and driving the most IOPS.

Balance > Reports > Standard Reports > Server Reports > Servers With Misaligned Partitions

Use Case 8: Use the Performance Index to Identify Virtual Hosts That Need Optimization

Balance > Servers > Virtual Hosts
View the entire list and look at the Performance Index (PI) values. Hosts with a PI > 125 are currently experiencing degraded performance - they are either over-utilized with too many VMs, or under-provisioned for their current workload. Hosts with a PI well below 75 are not being sufficiently utilized and are wasting datacenter resources. As a general rule, strive to have all virtual hosts operating with PI values between 75 and 125, this provides the best possible balance between resource utilization and performance.

Balance > Servers > Virtual Hosts > select a host > PI tab
Balance > Reports > Standard Reports > Server Reports > Virtual Host Server Performance Index
Balance > Reports > Reports Scheduling > (to schedule the report on a regular basis)

Use Case 9: Display a Baseline View of Your Storage Environment

Balance > Reports > Scorecard Reports > Storage Scorecard
The Storage Scorecard report provides a baseline view of your storage environment with notable columns:
Percent Capacity - watch for anything > 75% used.
Disk Utilization - watch for anything averaging > 60%, or maximum > 75%
Also, watch for abnormally high response times, IOPS, or read and write throughput values (additional workloads could degrade the performance of all servers that share the storage!)

Notable Reports

Among all the excellent reports available in Balance, the below are highly recommended for having a look at, and all (except the Servers With Misaligned Partitions - which you are probably not going to look at that often) are worth putting on a schedule too:
- Application Storage Trends
- Array Utilization
- Server Volume Capacity Forecast
- Servers with Misaligned Partitions
- Storage Scorecard
- Virtual Host Server Performance Index
- Virtual Machine Scorecard