The PowerMax storage operating system, PowerMaxOS, uses the service level that is associated with each storage group to maintain system performance. Each service level corresponds to a target response time, which is the average response time expected for the storage group based on the selected service level.
PowerMaxOS defines the following service levels:
In our use case scenarios, we assigned different service levels to achieve the essential range of performance for the OLTP and DSS workloads:
The following table shows the service levels that are associated with the three use case tests that we conducted:
Table 6. PowerMaxOS service levels implemented in the three use cases
Use case |
PowerMax service level |
OLTP workload |
Diamond |
DSS workload |
Bronze |
Snapshot OLTP workload |
Diamond |
We used Dell EMC Live Optics for collecting data and validating the use case tests that are described in this document. Live Optics is a free, agentless software used for collecting data from PowerEdge servers. In just minutes, any user can set up Live Optics to collect a wealth of information for configuration and resource utilization analysis. The intuitive Live Optics dashboard enables DBAs to monitor and collect data across the server and VMware virtualization layers.
The following figure shows a Live Optics dashboard. The left side shows performance at the project, hypervisor, virtual server, and shared disk levels. The right side shows the collected data and graphs that can assist with quick visual analysis.
Figure 2. Live Optics dashboard
We used Live Optics during the validation testing to gather the data shown in tables and graphs throughout this guide. The following table lists the data source for each performance metric in our validation tests.
Table 7. Performance metric sources
Performance metric |
Source (report) |
CPU utilization |
Dstat |
TPM |
HammerDB |
NOPM |
HammerDB |
IOPS |
Unisphere |
Storage latency in milliseconds |
Unisphere |
Throughput in megabytes per second |
Unisphere |