HP StorageWorks SAN Virtualization Services Platform Administrator Guide (5697-0934, May 2011)

4 Monitoring the SVSP domain
This chapter describes how to set up monitoring for an SVSP domain using administrative tools.
Array workload concentration
SVSP relies on the back-end arrays to handle the I/O workload. The volume management
capabilities permit focusing the workload of multiple front-end virtual disks onto one back-end
virtual disk. The DPMs can unintentionally concentrate front-end I/O workload from multiple front-end
hosts and front-end paths down a single back-end path. Careful design and monitoring are required
to ensure that the arrays are not running too close to saturation. Following the storage pool
configuration best practices (see “Building basic storage pools” (page 97)) is the first step in
avoiding array workload concentration resulting in array saturation. Monitoring the array
performance is the second step. When an array is found to be operating too close to saturation
(as defined by the array manufacturer), SVSP data migration can be used to migrate virtual disks
off of the overloaded array and onto an array with spare capacity and performance. Array-based
tools, like EVAperf for the EVA, should be used to monitor the load on back-end physical LUs.
Unless specifically conducting stress testing, storage should not be run “in the Red-Zone” and should
be 80% or less loaded during normal testing. If maintaining performance is critical during a
component failure, this number may be closer to 40%.
Monitoring system performance
There are several ways to monitor SVSP performance:
• Monitor the system with the VSM GUI on a regular basis.
• Monitor the application server-to-array throughput using the Fibre Channel switch vendor
performance tools.
• Monitor the internal VSM data moving performance using a tool for collecting performance
data like Microsoft's Perfmon, which is described below.
• Monitor performance at the array using array-supplied tools.
• Monitor cross sectional bandwidth. This is the sum of the bandwidth of all links between the
servers and the DPMs, or the DPMs and the arrays.
System health monitoring
To help make sure that the VSM and the DPM system has maximum system uptime, it is important
to monitor the system health on ongoing basis using these tools:
• VSM graphical user interface
• VSM event log
• Alerts
• DPM SNMP traps
VSM graphical user interface
For best maintenance performance, open the VSM GUI on a daily basis and look for changes to
the objects status. Object status should be Normal (for logical objects like volumes or pools) or
Present (for physical devices like disk drives or HBAs). Figure 19 (page 44) shows an VSM GUI
in which the logical objects have a status of Normal and the physical devices have a status of
Present. Alternatively, you can use a Query to check for statuses that are not normal or not present.
Array workload concentration 43