Appendix D: Webbench Test Methodology - HP ML570 - ProLiant - G2 Tuning Manual

Novell NetWare 6 performance tuning guidelines for ProLiant servers
appendix d:
WebBench test
methodology
The Ziff-Davis WebBench 4.1 test (NetWare_simple_nsapi_wb41) was used to measure
the performance of the Web server software and hardware. For additional information on
WebBench, visit www.webbench.com/benchmarks/webbench/webbench.asp?visitor=X.
The dynamic version (NetWare_simple_nsapi_wb41) was executed on sixty clients with
minor modifications. For more details, refer to the notes listed later in this section of the
document.
The LOS on all clients was Windows 2000 (SP2). The clients were connected to the two
switches through 100 MB/s full duplex connections to two Anritsu multiflow 5048
switches. The two switches were configured as two separate LAN segments (30 clients per
LAN segment). Next, the ProLiant ML570 G2 server was connected to each of the switches
with a CAT5 cable to form the two LAN segments. Refer to server and test-bed tables for
detailed information about the server, client, and the test-bed.
The executions of the slightly modified WebBench standard test suites start with a single
client issuing requests and increments the client workload by four until the number of clients
issuing the requests reaches sixty.
Notes:
The number of threads in simple_nsapi_wb41.tst script was increased from 1 thread
to 5 threads to put more workload stress on the server with the least amount of clients.
The workload receive buffer size was changed from the default 4 K to 32 K for the
test.
The HTTP configuration for NetWare_simple_nsapi_wb41 test was modified as
follows:
%HPPT1.0 request 0
% persistent connection requests 100
minimum request per PC 20
maximum request per PC 30
% pipeline request 0
47