HP StorageWorks Enterprise Virtual Array Online Firmware Upgrade Best Practices (5697-6388, January 2007)
Table 4 Application/Database layout
Microsoft Exchange Server 2003 SP2
Number of users (light)
750
Number of users (heavy)
2000
Load Profile
MMB3
Number of Storage Groups
2
Number of Databases per Storage Group
2
Mailbox Size
100MB
Load Simulation Tool LoadSim 2003
Microsoft SQL Server 2005
Transaction Ty
pe
TPC-C OLTP
Read/W rite Distribution 70/30
Database Size
120GB
Client Connec
tions (light)
50
Client Connections (heavy)
200
Load Simulation Tool Benchmark Factory
Oracle 10g
Transaction Type
TPC-C OLTP
Read/W rite Distribution 70/30
Database Size
120GB
Client Connections (light)
50
Client Connections (heavy)
200
Load Simula
tion Tool
Benchmark F
actory
Test cases
Baseline t
est
Performed baseline test of the online co de loa d operation. The objective was to test the firmware
upgrade and determine the length of time the upgrade will take. The expected outcome was a successful
non-disruptive upgrade of the EVA firmware.
The test steps included the following:
1. Perform non-disruptive firmware upgrade.
2. Capture firmware upgrade duration.
3. Perform firmware downgrade.
The data to be captured was the duration of firmware upgrade.
28
Effect of online firmware upgrade on application resiliency