Tests Results With Http Only - Compaq 117755-003 - ProSignia - 740 White Paper

Performance analysis and tuning of raptor’s eagle nt 3.06 firewall on compaq servers
Hide thumbs Also See for 117755-003 - ProSignia - 740:
Table of Contents

Advertisement

278A/0497
W
P
HITE
APER
(cont.)
. .
. .
. .
Other Systems and Configurations
. .
. .
. .
This section includes test runs with other systems and configurations. It also includes other runs
. .
using the base system described above with other configurations.
. .
. .
Graph 7 includes runs #16, #17, and #18 for ProSignia 500, ProLiant 800, and the ProLiant 4500
. .
. .
respectively. These runs should not be compared to the base system because of the configuration
. .
and hardware differences. Descriptions of these three machines and their configurations are listed
. .
. .

in the
section Test Configurations of the Firewall.
. .
. .
. .
. .
. .
. .
1100
. .
. .
1000
. .
900
. .
. .
800
. .
700
. .
600
. .
. .
500
. .
400
. .
. .
300
. .
200
. .
100
. .
. .
. .
. .
. .
. .
. .
. .
. .
. .
Graph 7: ProSignia 500, ProLiant 800, ProLiant 4500
. .
. .
The low-end server, the ProSignia 500, had an overall average of 273 TPM for all virtual clients
. .
pushing almost an estimated 1.9Mb per seconds through the system. The ProLiant 800 had an
. .
. .
overall average for all virtual clients of 620 TPM with a high of 665 TPM at 48 virtual clients.
. .
The ProLiant 4500 had an overall average of 398 TPM for its virtual client runs. The average
. .
. .
failure rates for these test runs are less than 1%.
. .
. .
Another test run on the base system showed that the firewall could handle more than 10Mb of
. .
. .
traffic through the firewall for HTTP and FTP transfers. This test added the number of virtual
. .
clients to a total of 144 virtual clients on 8 physical client machines. The test results show that the
. .
. .
TPM was 1696 and 10.67Mb throughput was calculated for data passing through the firewall.
. .
. .
. .

Tests Results with HTTP Only

. .
. .
The same model is used for HTTP only sets of tests as for the tests with HTTP and FTP. The
. .
. .
same base system is used with no changes except that 100 percent of the traffic is HTTP traffic.
. .
The benchmark is run using the base system with the same test bed setup, 3 physical servers and
. .
. .
8 physical clients, with 1 to 72 virtual clients and 6 virtual servers. The set of tests run in this
. .
section were designed to show the performance increases that had been made to the HTTP
. .
. .
daemon. These increases in performance from the HTTP daemon are attributed to the threaded
. .
enhancement, the use of IO completion ports, and caching of the rules. This section presents the
. .
. .
Base Run, DNS Lookups for HTTPD and HTTP Cache, and 100 Rules.
. .
. .
. .
. .
. .
. .
.
20
Runs #16, #17, and #18 in TPM
0
1
12
24
32 36 48
Virtual Clients
Prosignia 500,
P120
Proliant 800, PP200,
256C
Proliant 4500, P133
56 72

Advertisement

Table of Contents
loading

Table of Contents