HP StorageWorks Enterprise Virtual Array Online Firmware Upgrade Best Practices (5697-6388, January 2007)

Light load te
sting
Performed mu
ltiple online code loads under a light load. The objective was to observe application
resiliency t
othermware upgrade. The expected outcome was that the rmware upgrade should
complete wi
thout the applications failing or having excessive latency.
The test ste
ps included the following:
1. Start load simulators on all three applications.
2. Start capturing metrics with Windows Performance Monitor.
3. Allow I/O to run for at least one hour.
4. Perform
rmware upgrade.
5. Save captured metrics.
6. Repeat steps 1-5 two more times.
Heavy load testing
Performed multiple online code loads under a heavy load. The objective was to observe application
resiliency to the rmware upgrade. The expected outcome was that the rmware upgrade should
complete without the applications failing or h aving excessive latency.
The test steps included the following:
1. Start load simulators on all three applications.
2. Start capturing metrics with Per fmon.
3. Allow I/O to run for at least one hour.
4. Perform rmware upgrade.
5. Save captured metrics.
6. Repeat steps 1-5 two more times.
HP StorageWorks Enterprise Virtual Array online rmware upgrade best practices
29