IBM Storage Virtualize guideline values for key performance indicators

To improve the performance and resiliency of your storage environment, compare the guideline values for key performance indicators with the values reported for your storage systems and devices.

Guideline values for performance were established by monitoring, measuring, analyzing, and stress testing the performance of IBM Storage Virtualize storage systems.

The name IBM Storage Virtualize is used to refer to the following types of storage systems:
  • IBM® SAN Volume Controller
  • IBM Storage Virtualize for Public Cloud
  • IBM Storage Virtualize as Software Only
  • IBM Storwize® block storage systems
  • IBM FlashSystem® devices that run IBM Storage Virtualize

Try it out! From the Resources menu, click Block Storage Systems. Double-click a storage system that runs IBM Storage Virtualize, and click Performance in the General section of the navigation pane.

Performance metrics collected over 24 hours

By default, the charts compare the performance metrics that were collected each hour over the last 24 hours with the guideline values. You can use the calendar, which is next to the title of the page, to change the date and compare current values with historical values.

In most of the charts, a horizontal line is used to indicate the guideline value for the metric. If your devices report prolonged periods of slow response times, you can check whether the performance values for your devices are over the guideline values.

For example, you experience slow send response times for a node in a cluster. You check the chart that tracks the send response times for the nodes and see that one of the node's response times is higher than the guideline value. You can then take remedial action such as balancing the workload of the nodes across the cluster. Alternatively, you can move some of the workloads to other storage systems.

The following key performance indicators are analyzed:

Max Cache Fullness by Pool
The maximum amount of the lower cache that the write cache partitions on the nodes that manage the pool are using for write operations. If the value is 100%, one or more cache partitions on one or more pools is full. The operations that pass through the pools with full cache partitions are queued and I/O response times will increase for the volumes in the affected pools. Available in IBM Storage Virtualize 7.3 or later.
Guideline value The guideline value is 80%.
Alert A critical alert for max write cache fullness is automatically generated when the value is equal to or more than 99%.
Overall Port Bandwidth Percentage by Port
The percentage of the port bandwidth that is used for receive and send operations. This value is an indicator of port bandwidth usage that is based on the speed of the port.
Guideline value The guideline value is 50%.
Compare the guideline value for this metric with the values measured for the switch ports. Because a cluster can have many ports, the chart shows only the eight ports with the highest average bandwidth.
Alert A warning alert is automatically generated when the value for port receive bandwidth or port send bandwidth is equal to or more than 75%. A critical alert is generated when the value for port receive bandwidth or port send bandwidth is equal to or more than 85%.
Port-to-Local Node Send Response Time by Node
The average number of milliseconds to complete a send operation to another node that is in the local cluster. This value represents the external response time of the transfers.
Guideline value The guideline value is 0.6 ms/op.
Port-to-Remote Node Send Response Time by Node
The average number of milliseconds it takes to complete a send operation to a node in the remote cluster. And, the average number of milliseconds it takes to complete a receive operation from a node in the remote cluster. This value represents the external response time of the transfers.
A guideline value isn't available for this metric because response times for copy-service operations can vary widely.

You can compare the response times to identify discrepancies between the response times for the different nodes.

Read Response Time by I/O Group
The average number of milliseconds to complete a read operation.
Guideline value The guideline value is 15 ms/op.
Write Response Time by I/O Group
The average number of milliseconds to complete a write operation.
Guideline value The guideline value is 5 ms/op.
Node Utilization Percentage by Node
The average bandwidth percentages for the ports in the node that are actively used for host and MDisk receive and send operations. The average is weighted based on the port speed and the technology limitations of the node hardware.
Guideline value The guideline value is 60%.
Port Send Delay Time
The average number of milliseconds of delay that occurs on the port for each send operation. The reason for these delays might be a lack of buffer credits.
A guideline value isn't available for this metric because delay times can vary significantly depending on configuration and usage.
Compare the delay times to identify discrepancies between the ports' delay times and any spikes that might correlate with the time of any reported performance problems.
The Port Send Delay Time is shown instead of the Zero Buffer Credit Percentage by Node chart for some IBM Storage FlashSystem storage systems, such as IBM Storage FlashSystem 9110.
Zero Buffer Credit Percentage by Node

The amount of time, as a percentage, that the port wasn't able to send frames between ports because of insufficient buffer-to-buffer credit. The amount of time value is measured from the last time that the node was reset. In Fibre Channel (FC) technology, buffer-to-buffer credit is used to control the flow of frames between ports.

Information about zero buffer credit is only collected and analyzed for 8 Gbps FC ports.

Guideline value The guideline value is 20%.
Tip: When you add a storage system, a default alert policy is assigned to the storage system. For example, when you add IBM SAN Volume Controller, the default policy for IBM SAN Volume Controller is automatically assigned to the storage system. To find out which alerts are in the default policy, click the Configuration menu and click Alert Policies. To see the alerts that are defined for the policy, double-click the default policy.

More actions

Review the following resources to find out:

Restrictions

The following charts aren't available for storage systems that run IBM Storage Virtualize and use the iSCSI protocol to connect to storage systems:
  • Overall Port Bandwidth Percentage by Port
  • Node Utilization Percentage by Node
  • Port Send Delay Time