Report Archive for

Click on a report title to go to the download page for that report.

Radware Ltd. Radware AppDirector x08 and x16 Application Switches: Performance Evaluation versus F5 Networks BIG-IP 1600 and 3600 Document number: 210104 Release Date: 12 Apr 2010 Radware commissioned Tolly to validate the performance of the AppDirector 4016 and the newly-introduced 4008 vs F5’s newly-introduced BIG-IP 1600 and 3600 running F5’s newest version, 10.0. This comes two years after the release of the Tolly report showing AppDirector x16’s superior performance over the F5 BIG-IP 1500, 6400 and 6800 platforms. The project measured the throughput, transaction rates and response times of the systems while processing layer 7 transactions with and without SSL termination. Tests also measured system performance while the systems were subject to a number of denial-of-service attacks. The test methodology used for this report relies upon test procedures, metrics and documentation practices as defined by Tolly Common Test Plan #1077 Application Switch Performance. Application Switches
Radware Ltd. Radware AppDirector x08 and x16 Application Switches: Performance Evaluation versus F5 Networks BIG-IP 1600 and 3600 Document number: 210104JP Release Date: 17 Mar 2010 Radware commissioned Tolly to validate the performance of the AppDirector 4016 and the newly-introduced 4008 vs F5’s newly-introduced BIG-IP 1600 and 3600 running F5’s newest version, 10.0. This comes two years after the release of the Tolly report showing AppDirector x16’s superior performance over the F5 BIG-IP 1500, 6400 and 6800 platforms. The project measured the throughput, transaction rates and response times of the systems while processing layer 7 transactions with and without SSL termination. Tests also measured system performance while the systems were subject to a number of denial-of-service attacks. The test methodology used for this report relies upon test procedures, metrics and documentation practices as defined by Tolly Common Test Plan #1077 Application Switch Performance. Application Switches
Brocade Communications Systems, Inc. Brocade ServerIron ADX 1000 Application Switch with Secure Sockets Layer (SSL) Acceleration Document number: 210101 Release Date: 20 Jan 2010 Brocade Communications, Inc. commissioned Tolly to validate the SSL and layer 7 performance of its ServerIron ADX 1000 application switch. Test measured both transaction and data throughput for SSL (encrypted) and unencrypted layer 7 traffic. Results were compared with vendor-published results for similar products from Cisco Systems, Citrix Systems and F5 Networks. Cost per transaction and cost per Gigabit per second of throughput were calculated for all of the systems. Tests were based on the Tolly Common Test Plan #1077 Application Switch Performance. (For more information see http://CommonTestPlan.org.) Application Switches
Brocade Communications Systems, Inc. Brocade ServerIron ADX 10000 10 GbE Performance and Scalability Document number: 209150 Release Date: 01 Nov 2009 Brocade Communications, Inc. commissioned Tolly to validate the performance of its ServerIron ADX 10000 application switch. This device was outfitted with 16 10 Gigabit Ethernet connections and ran 32 application processing cores. Test covered L4-7 performance measuring transactions and connections per second, aggregate bandwidth, DNS query response capacity and resistance to distributed denial of service (DDoS) attacks. Application Switches
Citrix Systems, Inc. Citrix NetScaler MPX 17000 v9.1 Application Delivery Controller with nCore Technology: Latency versus F5 BIG-IP v10 Document number: 209127 Release Date: 30 Jun 2009 Citrix Systems commissioned Tolly to benchmark the latency of its NetScaler MPX17000 application delivery controller and compare that to the F5 BIG-IP 8800. Tests were run using the latest software from both vendors with each appliance outfitted with its maximum number of 10GbE interfaces. Ixia network testing equipment was used to drive various loads ranging from 1,200 to 1.2 million concurrent connections. The Citrix NetScaler appliance consistently delivered significantly lower latency than the F5 BIG-IP across the range of tests. Click on the report icon to purchase the report. Application Switches
Brocade Communications Systems, Inc. Brocade ServerIron ADX 8000 10 GbE Performance and Scalability Document number: 209122 Release Date: 11 May 2009 Brocade Communications, Inc. commissioned Tolly to validate the performance of its ServerIron ADX 8000 application switch. This device was outfitted with 16 10 Gigabit Ethernet connections and ran 32 application processing cores. Test covered L4-7 performance measuring transactions and connections per second, aggregate bandwidth, DNS query response capacity and resistance to distributed denial of service (DDoS) attacks. This is the FIRST test to be published in accordance with Tolly Common Test Plan #1077 Application Switches V1.1. That document is available for license and you can find more information about it on the Tolly Common Test Plan website. Note: The actual test environment was demonstrated at the Brocade booth at Interop Las Vegas May 2009. Download the free test report. Application Switches
Radware Ltd. "Talking Outside the Box": Podcast Interview with Radware's Nir Ilani and Overview of Tolly Testing Document number: 208333 Release Date: 05 Dec 2008 This 12-minute podcast focuses on Radware's series of OnDemand Switches. Tolly Group Founder, Kevin Tolly, provides an overview of a recent Tolly Group tests of a number of Radware switches. In addition, Charlie Bruno, Executive Editor of The Tolly Group, discusses the ramifications of the test with Nir Ilani, Product Marketing Director at Radware. Details of the test can be found in Tolly Group document 208285 through 208288. Application Switches
Radware Ltd. Radware OnDemand Switch 1 & OnDemand Switch 2 AppDirector Version 1.06 Competitive Performance Evaluation versus F5 Networks BIG-IP 6800 Document number: 208288 Release Date: 16 Sep 2008 Radware commissioned The Tolly Group to evaluate the performance of its OnDemand Switch 1 & 2 AppDirector, the company’s nextgeneration application switch, in comparison with F5 Network’s BIG-IP 6800. The goal was to measure performance while handling challenging tasks to determine which device offers the higher Layer 7 transaction rate and faster response time. Tolly Group engineers determined the transactions-per-second (tps) rate and the corresponding throughput and response time for 10 object sizes at Layer 7 in multiple scenarios of a single HTTP request per connection and 10 HTTP requests per connection. They also in conjunction tested the ability of the platforms to maintain performance while dealing with DDoS attack packets. Tests were conducted in February 2008. Download the free report. Application Switches
Radware Ltd. Radware OnDemand Switch 1 & OnDemand Switch 2 AppDirector Version 1.06 Competitive Performance Evaluation versus F5 Networks BIG-IP 6400 Document number: 208287 Release Date: 16 Sep 2008 Radware commissioned The Tolly Group to evaluate the performance of its OnDemand Switch 1 & 2 AppDirector, the company’s nextgeneration application switch, in comparison with F5 Network’s BIG-IP 6400. The goal was to measure performance while handling challenging tasks to determine which device offers the higher Layer 7 transaction rate and faster response time. Tolly Group engineers determined the transactions-per-second (tps) rate and the corresponding throughput and response time for 10 object sizes at Layer 7 in multiple scenarios of a single HTTP request per connection and 10 HTTP requests per connection. They also in conjunction tested the ability of the platforms to maintain performance while dealing with DDoS attack packets. Tests were conducted in February 2008. Download the free report. Application Switches
Coyote Point Systems, Inc. Coyote Point Systems Inc. Equalizer E350si Competitive Performance and Feature Evaluation versus Barracuda Networks Load Balancer 340 Document number: 208329 Release Date: 05 Sep 2008 Coyote Point Systems Inc. commissioned The Tolly Group to evaluate the Equalizer E350si, a Layer 4-7 load-balancing appliance designed to provide Enterprise-class performance aimed at small- to medium-sized businesses (and budgets). Coyote Point also asked The Tolly Group to evaluate the Equalizer E350si against Barracuda Networks Load Balancer 340. The Barracuda Load Balancer 340 was chosen because the vendor’s specifications show it offers similar capabilities to the Equalizer E350si. Tolly Group engineers measured the transaction rate, response time and aggregate throughput at Layer 4 and Layer 7 when handling 64-byte and 16Kbyte object sizes. Engineers also verified the features from various categories such as scalability, flexibility and ease of management. Tests were conducted in July 2008. Application Switches
Radware Ltd. Radware OnDemand Switch 1 AppDirector Version 1.06 Competitive Performance Evaluation versus F5 Networks BIG-IP 1500 Document number: 208286 Release Date: 31 Jul 2008 Radware commissioned The Tolly Group to evaluate the performance of its OnDemand Switch 1 AppDirector Version 1.06, the company’s next-generation platform, in comparison with F5 Network’s BIG-IP 1500. Tolly Group engineers determined the connection-per-second (cps) rate and the corresponding throughput and response time for 10 object sizes at Layer 7 in multiple scenarios of a single HTTP request per connection and 10 HTTP requests per connection, while also testing the ability of the platforms to maintain service levels while dealing with DDoS attack. Tests were conducted in February 2008. Application Switches
A10 Networks AX 3200 Ver. 1.2.3 Performance Evaluation of AX 3200 Advanced Traffic Manager with HTTP Caching Document number: 208312 Release Date: 05 May 2008 A10 Networks, Inc. commissioned The Tolly Group to evaluate the AX 3200 Advanced Traffic Manager, A10’s next generation server load balancer. A10 Networks asked The Tolly Group to test the AX 3200 using the same test methodologies used by F5 Networks in its public performance report. In January 2008, Tolly Group engineers benchmarked the connection-per-second (cps) rate, transaction-per-second (tps) rate, SYN cookies, and aggregate throughput of the AX 3200 at Layer 4, Layer 7 plus SSL sessions for five object sizes ranging from 128- to 512K-bytes and for different numbers of HTTP requests per TCP connections. In April 2008, engineers also measured the Layer 7 switching performance with HTTP Caching enabled. Application Switches
A10 Networks VirtualN Technology Performance and Scalability Evaluation of VirtualN Running on AX 3200 Advanced Traffic Manager Document number: 208295 Release Date: 29 Apr 2008 A10 Networks, Inc. commissioned The Tolly Group to evaluate the company’s VirtualN technology, an innovative Web Application Delivery on-demand solution running on multiple AX 3200 Advanced Traffic Manager devices. Tolly Group engineers benchmarked the connection-per-second (cps) rate, transaction-per-second (tps) rate, SSL capability, HTTP caching and aggregate throughput of one, two and four units of the AX 3200 at Layer 4 and Layer 7 for two object sizes, 128 bytes and 8 Kbytes. Tests were conducted in April 2008. Application Switches
Radware Ltd. OnDemand Switch 1 & OnDemand Switch 2 AppDirector Version 1.06 Competitive Performance Evaluation versus F5 Networks BIG-IP 6800, BIG-IP 6400 and BIG-IP 1500 Document number: 208285 Release Date: 01 Apr 2008 Radware commissioned The Tolly Group to evaluate the performance of its OnDemand Switches, the vendor’s next-generation platforms that reportedly delivers over 3.7 Gbps. Engineers valuated the performance of the OnDemand switches against F5 Network’s BIG-IP 6800, BIG-IP 6400 and BIG-IP 1500 platforms. The goal was to measure performance while also handling hallenging tasks, Layer 7 operations and transaction processing under DDoS attacks, and validate the readiness of the products for next-generation applications. Engineers measured the transaction-persecond (tps) rate, throughput and response time for 10 object sizes at Layer 7 in multiple scenarios of a single HTTP request per connection and 10 HTTP requests per connection, while also testing the ability of the platforms to maintain performance while dealing with DDoS attack packets. Tests were conducted in February 2008. Application Switches
A10 Networks A10 Networks, Inc. AX 3200 Ver. 1.2 Performance Evaluation of Application Delivery Controller Document number: 208280 Release Date: 23 Jan 2008 A10 Networks, Inc. commissioned The Tolly Group to evaluate the AX 3200 Advanced Traffic Manager, A10’s next-generation server load balancer. A10 Networks asked The Tolly Group to test the AX 3200 using the same test cases used by F5 Networks in its publicly available performance report. Additionally, price/performance calculations were performed. Tolly Group engineers benchmarked the connection-per-second (cps) rate, transaction-per-second (tps) rate, SYN cookies, and aggregate throughput of the AX 3200 at Layer 4, Layer 7 plus SSL sessions for five object sizes ranging from 128- to 512K-bytes. Then, engineers calculated the price/performance. Tests were conducted in December 2007. Application Switches
A10 Networks A10 Networks, Inc. AX 2100 Ver. 1.2 Competitive Performance Evaluation versus F5 Networks BIG-IP 3400 ver. 9.2.5 Document number: 208265 Release Date: 23 Jan 2008 A10 Networks, Inc. commissioned The Tolly Group to evaluate the price/performance advantages of the AX 2100 Advanced Traffic Manager, a next-generation server load balancer that serves up Enterprise-class performance in a cost-effective fashion. A10 Networks asked The Tolly Group to use the same test cases as an F5 Networks publicly available performance report and evaluate the AX 2100 alongside F5 Networks’ BIG-IP 3400. Tolly Group engineers determined the connection-per-second (cps) rate at Layer 4, the SSL transaction-per-second (tps) rate at Layer 7 and aggregate throughput of the AX 2100 and the F5 BIG-IP 3400. Engineers used the performance results to calculate price/performance metrics. Testing was conducted in December 2007. Application Switches
Nortel Nortel Application Switch 2424/3408 Evaluation of Security, Resiliency and Load Balancing on a Network Supporting Microsoft Live Communications Server 2005 and Nortel Multimedia Communications Server 5100 Document number: 207179 Release Date: 16 Feb 2007 Nortel commissioned The Tolly Group to evaluate a converged network infrastructure for Enterprise users to illustrate the load-balancing capabilities of its Nortel Application Switch 2424/3408. Tests attempted to show Nortel Application Switch support for Server Load Balancing for servers running Nortel MCS or Microsoft LCS - either locally, or across the network. Tests also illustrated the ability of the Nortel Application Switch to detect and block security attacks like Distributed Denial of Service (DDoS) attacks. Tests were conducted in November 2006. Application Switches
Nortel Nortel Application Switch 2424/3408 Evaluation of Resiliency and Load Balancing on a Network Supporting Microsoft Live Communications Server 2005 Document number: 207177 Release Date: 17 Jan 2007 Nortel commissioned The Tolly Group to evaluate a converged network infrastructure for Enterprise users to illustrate the load balancing capabilities of its Nortel Application Switch 2424/3408. The series of tests called for demonstrating that Application Switches can load balance user sessions across Microsoft Live Communications Server 2005 (LCS Server) within a pool, both on–site and across a network. Tests also aimed to verify that the Application Switch provides resiliency by seamlessly taking over the available and new user sessions from other sites in the network, in the event of a failure of an Application Switch or LCS server. Application Switches
Coyote Point Systems, Inc. Coyote Point Systems Equalizer E350si Competitive Performance Evaluation versus F5 Networks BIG-IP 1500 and BIG-IP 3400 Document number: 205131 Release Date: 14 Nov 2005 Coyote Point Systems, Inc. commissioned The Tolly Group to evaluate the Equalizer E350si, an entry-level Layer 4/7 load-balancing switch designed to serve up Enterprise-class performance in a cost effective fashion. Furthermore, Coyote Point asked The Tolly Group to evaluate the Equalizer E350si solution alongside several comparable offerings from F5 Networks. Tolly Group engineers tested the Equalizer E350si, measuring its Layer 4 connection rate, Layer 7 transaction rate and aggregate throughput (Mbps) at both Layer 4 and Layer 7. Engineers benchmarked the performance of the Equalizer E350si against an F5 Networks BIG-IP 1500 and a BIG-IP 3400, two devices that cost 2.5X to almost 4X more than the Equalizer E350si. Tests were conducted by Tolly Group engineers in September 2005 at Tolly Group headquarters in Boca Raton, FL. Results show that the Equalizer E350si outperformed the BIG-IP 1500 in every test, offered near-comparable performance to the much more expensive BIG-IP 3400, and offered compelling price/performance advantages. Application Switches

Radware Ltd. Radware AppDirector x08 and x16 Application Switches: Performance Evaluation versus F5 Networks BIG-IP 1600 and 3600

Document number: 210104
Release Date: 12 Apr 2010

Radware commissioned Tolly to validate the performance of the AppDirector 4016 and the newly-introduced 4008 vs F5’s newly-introduced BIG-IP 1600 and 3600 running F5’s newest version, 10.0. This comes two years after the release of the Tolly report showing AppDirector x16’s superior performance over the F5 BIG-IP 1500, 6400 and 6800 platforms.

The project measured the throughput, transaction rates and response times of the systems while processing layer 7 transactions with and without SSL termination. Tests also measured system performance while the systems were subject to a number of denial-of-service attacks.

The test methodology used for this report relies upon test procedures, metrics and documentation practices as defined by Tolly Common Test Plan #1077 Application Switch Performance. Application Switches

Radware Ltd. Radware AppDirector x08 and x16 Application Switches: Performance Evaluation versus F5 Networks BIG-IP 1600 and 3600

Document number: 210104JP
Release Date: 17 Mar 2010

Radware commissioned Tolly to validate the performance of the AppDirector 4016 and the newly-introduced 4008 vs F5’s newly-introduced BIG-IP 1600 and 3600 running F5’s newest version, 10.0. This comes two years after the release of the Tolly report showing AppDirector x16’s superior performance over the F5 BIG-IP 1500, 6400 and 6800 platforms.

The project measured the throughput, transaction rates and response times of the systems while processing layer 7 transactions with and without SSL termination. Tests also measured system performance while the systems were subject to a number of denial-of-service attacks.

The test methodology used for this report relies upon test procedures, metrics and documentation practices as defined by Tolly Common Test Plan #1077 Application Switch Performance. Application Switches

Brocade Communications Systems, Inc. Brocade ServerIron ADX 1000 Application Switch with Secure Sockets Layer (SSL) Acceleration

Document number: 210101
Release Date: 20 Jan 2010

Brocade Communications, Inc. commissioned Tolly to validate the SSL and layer 7 performance of its ServerIron ADX 1000 application switch.

Test measured both transaction and data throughput for SSL (encrypted) and unencrypted layer 7 traffic. Results were compared with vendor-published results for similar products from Cisco Systems, Citrix Systems and F5 Networks. Cost per transaction and cost per Gigabit per second of throughput were calculated for all of the systems.

Tests were based on the Tolly Common Test Plan #1077 Application Switch Performance. (For more information see http://CommonTestPlan.org.) Application Switches

Brocade Communications Systems, Inc. Brocade ServerIron ADX 10000 10 GbE Performance and Scalability

Document number: 209150
Release Date: 01 Nov 2009

Brocade Communications, Inc. commissioned Tolly to validate the performance of its ServerIron ADX 10000 application switch. This device was outfitted with 16 10 Gigabit Ethernet connections and ran 32 application processing cores.

Test covered L4-7 performance measuring transactions and connections per second, aggregate bandwidth, DNS query response capacity and resistance to distributed denial of service (DDoS) attacks. Application Switches

Citrix Systems, Inc. Citrix NetScaler MPX 17000 v9.1 Application Delivery Controller with nCore Technology: Latency versus F5 BIG-IP v10

Document number: 209127
Release Date: 30 Jun 2009

Citrix Systems commissioned Tolly to benchmark the latency of its NetScaler MPX17000 application delivery controller and compare that to the F5 BIG-IP 8800.

Tests were run using the latest software from both vendors with each appliance outfitted with its maximum number of 10GbE interfaces.

Ixia network testing equipment was used to drive various loads ranging from 1,200 to 1.2 million concurrent connections.

The Citrix NetScaler appliance consistently delivered significantly lower latency than the F5 BIG-IP across the range of tests.

Click on the report icon to purchase the report. Application Switches

Brocade Communications Systems, Inc. Brocade ServerIron ADX 8000 10 GbE Performance and Scalability

Document number: 209122
Release Date: 11 May 2009

Brocade Communications, Inc. commissioned Tolly to validate the performance of its ServerIron ADX 8000 application switch. This device was outfitted with 16 10 Gigabit Ethernet connections and ran 32 application processing cores.

Test covered L4-7 performance measuring transactions and connections per second, aggregate bandwidth, DNS query response capacity and resistance to distributed denial of service (DDoS) attacks.

This is the FIRST test to be published in accordance with Tolly Common Test Plan #1077 Application Switches V1.1. That document is available for license and you can find more information about it on the Tolly Common Test Plan website.

Note: The actual test environment was demonstrated at the Brocade booth at Interop Las Vegas May 2009.

Download the free test report. Application Switches

Radware Ltd. "Talking Outside the Box": Podcast Interview with Radware's Nir Ilani and Overview of Tolly Testing

Document number: 208333
Release Date: 05 Dec 2008

This 12-minute podcast focuses on Radware's series of OnDemand Switches. Tolly Group Founder, Kevin Tolly, provides an overview of a recent Tolly Group tests of a number of Radware switches.

In addition, Charlie Bruno, Executive Editor of The Tolly Group, discusses the ramifications of the test with Nir Ilani, Product Marketing Director at Radware. Details of the test can be found in Tolly Group document 208285 through 208288.

Application Switches

Radware Ltd. Radware OnDemand Switch 1 & OnDemand Switch 2 AppDirector Version 1.06 Competitive Performance Evaluation versus F5 Networks BIG-IP 6800

Document number: 208288
Release Date: 16 Sep 2008

Radware commissioned The Tolly Group to evaluate the performance of its OnDemand Switch 1 & 2 AppDirector, the company’s nextgeneration application switch, in comparison with F5 Network’s BIG-IP 6800.

The goal was to measure performance while handling challenging tasks to determine which device offers the higher Layer 7 transaction rate and faster response time. Tolly Group engineers determined the transactions-per-second (tps) rate and the corresponding throughput and response time for 10 object sizes at Layer 7 in multiple scenarios of a single HTTP request per connection and 10 HTTP requests per connection. They also in conjunction tested the ability of the platforms to maintain performance while dealing with DDoS attack packets.

Tests were conducted in February 2008.

Download the free report. Application Switches

Radware Ltd. Radware OnDemand Switch 1 & OnDemand Switch 2 AppDirector Version 1.06 Competitive Performance Evaluation versus F5 Networks BIG-IP 6400

Document number: 208287
Release Date: 16 Sep 2008

Radware commissioned The Tolly Group to evaluate the performance of its OnDemand Switch 1 & 2 AppDirector, the company’s nextgeneration application switch, in comparison with F5 Network’s BIG-IP 6400.

The goal was to measure performance while handling challenging tasks to determine which device offers the higher Layer 7 transaction rate and faster response time. Tolly Group engineers determined the transactions-per-second (tps) rate and the corresponding throughput and response time for 10 object sizes at Layer 7 in multiple scenarios of a single HTTP request per connection and 10 HTTP requests per connection. They also in conjunction tested the ability of the platforms to maintain performance while dealing with DDoS attack packets.

Tests were conducted in February 2008.

Download the free report. Application Switches

Coyote Point Systems, Inc. Coyote Point Systems Inc. Equalizer E350si Competitive Performance and Feature Evaluation versus Barracuda Networks Load Balancer 340

Document number: 208329
Release Date: 05 Sep 2008

Coyote Point Systems Inc. commissioned The Tolly Group to evaluate the Equalizer E350si, a Layer 4-7 load-balancing appliance designed to provide Enterprise-class performance aimed at small- to medium-sized businesses (and budgets). Coyote Point also asked The Tolly Group to evaluate the Equalizer E350si against Barracuda Networks Load Balancer 340. The Barracuda Load Balancer 340 was chosen because the vendor’s specifications show it offers similar capabilities to the Equalizer E350si.

Tolly Group engineers measured the transaction rate, response time and aggregate throughput at Layer 4 and Layer 7 when handling 64-byte and 16Kbyte object sizes. Engineers also verified the features from various categories such as scalability, flexibility and ease of management.

Tests were conducted in July 2008.

Application Switches

Radware Ltd. Radware OnDemand Switch 1 AppDirector Version 1.06 Competitive Performance Evaluation versus F5 Networks BIG-IP 1500

Document number: 208286
Release Date: 31 Jul 2008

Radware commissioned The Tolly Group to evaluate the performance of its OnDemand Switch 1 AppDirector Version 1.06, the company’s next-generation platform, in comparison with F5 Network’s BIG-IP 1500.

Tolly Group engineers determined the connection-per-second (cps) rate and the corresponding throughput and response time for 10 object sizes at Layer 7 in multiple scenarios of a single HTTP request per connection and 10 HTTP requests per connection, while also testing the ability of the platforms to maintain service levels while dealing with DDoS attack.

Tests were conducted in February 2008.

Application Switches

A10 Networks AX 3200 Ver. 1.2.3 Performance Evaluation of AX 3200 Advanced Traffic Manager with HTTP Caching

Document number: 208312
Release Date: 05 May 2008

A10 Networks, Inc. commissioned The Tolly Group to evaluate the AX 3200 Advanced Traffic Manager, A10’s next generation server load balancer. A10 Networks asked The Tolly Group to test the AX 3200 using the same test methodologies used by F5 Networks in its public performance report.

In January 2008, Tolly Group engineers benchmarked the connection-per-second (cps) rate, transaction-per-second (tps) rate, SYN cookies, and aggregate throughput of the AX 3200 at Layer 4, Layer 7 plus SSL sessions for five object sizes ranging from 128- to 512K-bytes and for different numbers of HTTP requests per TCP connections. In April 2008, engineers also measured the Layer 7 switching performance with HTTP Caching enabled.

Application Switches

A10 Networks VirtualN Technology Performance and Scalability Evaluation of VirtualN Running on AX 3200 Advanced Traffic Manager

Document number: 208295
Release Date: 29 Apr 2008

A10 Networks, Inc. commissioned The Tolly Group to evaluate the company’s VirtualN technology, an innovative Web Application Delivery on-demand solution running on multiple AX 3200 Advanced Traffic Manager devices.

Tolly Group engineers benchmarked the connection-per-second (cps) rate, transaction-per-second (tps) rate, SSL capability, HTTP caching and aggregate throughput of one, two and four units of the AX 3200 at Layer 4 and Layer 7 for two object sizes, 128 bytes and 8 Kbytes.

Tests were conducted in April 2008. Application Switches

Radware Ltd. OnDemand Switch 1 & OnDemand Switch 2 AppDirector Version 1.06 Competitive Performance Evaluation versus F5 Networks BIG-IP 6800, BIG-IP 6400 and BIG-IP 1500

Document number: 208285
Release Date: 01 Apr 2008

Radware commissioned The Tolly Group to evaluate the performance of its OnDemand Switches, the vendor’s next-generation platforms that reportedly delivers over 3.7 Gbps. Engineers valuated the performance of the OnDemand switches against F5 Network’s BIG-IP 6800, BIG-IP 6400 and BIG-IP 1500 platforms. The goal was to measure performance while also handling hallenging tasks, Layer 7 operations and transaction processing under DDoS attacks, and validate the readiness of the products for next-generation applications.

Engineers measured the transaction-persecond (tps) rate, throughput and response time for 10 object sizes at Layer 7 in multiple scenarios of a single HTTP request per connection and 10 HTTP requests per connection, while also testing the ability of the platforms to maintain performance while dealing with DDoS attack packets. Tests were conducted in February 2008.

Application Switches

A10 Networks A10 Networks, Inc. AX 3200 Ver. 1.2 Performance Evaluation of Application Delivery Controller

Document number: 208280
Release Date: 23 Jan 2008

A10 Networks, Inc. commissioned The Tolly Group to evaluate the AX 3200 Advanced Traffic Manager, A10’s next-generation server load balancer. A10 Networks asked The Tolly Group to test the AX 3200 using the same test cases used by F5 Networks in its publicly available performance report. Additionally, price/performance calculations were performed.

Tolly Group engineers benchmarked the connection-per-second (cps) rate, transaction-per-second (tps) rate, SYN cookies, and aggregate throughput of the AX 3200 at Layer 4, Layer 7 plus SSL sessions for five object sizes ranging from 128- to 512K-bytes. Then, engineers calculated the price/performance. Tests were conducted in December 2007.

Application Switches

A10 Networks A10 Networks, Inc. AX 2100 Ver. 1.2 Competitive Performance Evaluation versus F5 Networks BIG-IP 3400 ver. 9.2.5

Document number: 208265
Release Date: 23 Jan 2008

A10 Networks, Inc. commissioned The Tolly Group to evaluate the price/performance advantages of the AX 2100 Advanced Traffic Manager, a next-generation server load balancer that serves up Enterprise-class performance in a cost-effective fashion.

A10 Networks asked The Tolly Group to use the same test cases as an F5 Networks publicly available performance report and evaluate the AX 2100 alongside F5 Networks’ BIG-IP 3400.

Tolly Group engineers determined the connection-per-second (cps) rate at Layer 4, the SSL transaction-per-second (tps) rate at Layer 7 and aggregate throughput of the AX 2100 and the F5 BIG-IP 3400. Engineers used the performance results to calculate price/performance metrics. Testing was conducted in December 2007.

Application Switches

Nortel Nortel Application Switch 2424/3408 Evaluation of Security, Resiliency and Load Balancing on a Network Supporting Microsoft Live Communications Server 2005 and Nortel Multimedia Communications Server 5100

Document number: 207179
Release Date: 16 Feb 2007

Nortel commissioned The Tolly Group to evaluate a converged network infrastructure for Enterprise users to illustrate the load-balancing capabilities of its Nortel Application Switch 2424/3408.

Tests attempted to show Nortel Application Switch support for Server Load Balancing for servers running Nortel MCS or Microsoft LCS - either locally, or across the network. Tests also illustrated the ability of the Nortel Application Switch to detect and block security attacks like Distributed Denial of Service (DDoS) attacks. Tests were conducted in November 2006.

Application Switches

Nortel Nortel Application Switch 2424/3408 Evaluation of Resiliency and Load Balancing on a Network Supporting Microsoft Live Communications Server 2005

Document number: 207177
Release Date: 17 Jan 2007

Nortel commissioned The Tolly Group to evaluate a converged network infrastructure for Enterprise users to illustrate the load balancing capabilities of its Nortel Application Switch 2424/3408.

The series of tests called for demonstrating that Application Switches can load balance user sessions across Microsoft Live Communications Server 2005 (LCS Server) within a pool, both on–site and across a network. Tests also aimed to verify that the Application Switch provides resiliency by seamlessly taking over the available and new user sessions from other sites in the network, in the event of a failure of an Application Switch or LCS server.

Application Switches

Coyote Point Systems, Inc. Coyote Point Systems Equalizer E350si Competitive Performance Evaluation versus F5 Networks BIG-IP 1500 and BIG-IP 3400

Document number: 205131
Release Date: 14 Nov 2005

Coyote Point Systems, Inc. commissioned The Tolly Group to evaluate the Equalizer E350si, an entry-level Layer 4/7 load-balancing switch designed to serve up Enterprise-class performance in a cost effective fashion. Furthermore, Coyote Point asked The Tolly Group to evaluate the Equalizer E350si solution alongside several comparable offerings from F5 Networks.

Tolly Group engineers tested the Equalizer E350si, measuring its Layer 4 connection rate, Layer 7 transaction rate and aggregate throughput (Mbps) at both Layer 4 and Layer 7. Engineers benchmarked the performance of the Equalizer E350si against an F5 Networks BIG-IP 1500 and a BIG-IP 3400, two devices that cost 2.5X to almost 4X more than the Equalizer E350si.

Tests were conducted by Tolly Group engineers in September 2005 at Tolly Group headquarters in Boca Raton, FL. Results show that the Equalizer E350si outperformed the BIG-IP 1500 in every test, offered near-comparable performance to the much more expensive BIG-IP 3400, and offered compelling price/performance advantages. Application Switches