ZXTM (Zeus Extensible Traffic Manager) In Virtual Mode - Broadband Testing

Zeus ZXTM Virtual Mode – Microsoft Virtual Server ReportZXTM PUT TO THE TEST IN A VIRTUALISEDENVIRONMENTTestbed DetailsBroadband-Testing created a twin testbed in order to put ZXTM to the test in a virtualisedenvironment, a Microsoft Virtual Server platform.For the testing we were using MS Virtual Server 2005 R2, running on WindowsServer 2003 R2, on a dual AMD Opteron platform, typical of a standard ZXTMappliance environment.Benchmarking wise, we used both Zeus’ BenchBot, Linux-based web client simulators anda testbed based upon Spirent’s Avalanche and Reflector 2700 client and server trafficsimulator appliances (see appendix for more details).Figure 3 – Our Testbed For The Virtualisation TestingAs always, the basis of the testing was not to simply run a series of “white gloves on”technical laboratory tests, but to create as realistic as possible a virtual world for theZXTM, with real web traffic and real-world applications. It was our goal to define tests ina very clear way that use parameters that would be very meaningful to a customer foruse as a sizing-guide in a purchasing decision. It's not uncommon to see a vendor claim"connections per second" by measuring "null connections", or by only measuring the newconnections established (i.e. SYN's per second, not full connections where there is anactual transfer of data) in order to boost metrics. All the tests performed were carryingout real end user transactions, which is why we are able to state that these were realworld conditions. The scenarios we created and tested, were as follows:Requests Per Second (Layer 4 and Layer 7 - http)Maximum SSL Sessions (termination handled by ZXTM)L4/L7 ThroughputCache RPS and Throughput4 Broadband-Testing 1995-2007

Zeus ZXTM Virtual Mode – Microsoft Virtual Server ReportNote: For more reports on ZXTM, please go online to the Broadband-Testing website:www.broadband-testing.co.uk where you will find several reports available to downloadfor free.One important point to make, in terms of analysing these results against previous ZXTMtesting we have carried out, is that – in native form – ZXTM runs as a 64-bit application.However, in virtual mode it is only able to run as a 32-bit application, so there is anobvious potential performance loss point, in addition to the extra overhead involved in thelayered approach that virtualisation creates. It is worth noting again, too, that MS VirtualServer can only utilise a single CPU.Test: Requests Per Second (RPS) At Layer 4 And Layer 7Here we were looking for a measure of how many concurrent connections – effectiveInternet sessions - ZXTM could support during an extended period of user activity. Thisnot only gives us an indication of the scalability – per device – of the ZXTM, but also itsresilience, vital within a virtualised environment. Specifically, we looked to see how manyrequests per second each device was capable of sustaining across a range of file requestsizes. Max requests is with a minimal file size, then we tested again with 2K and 8K filesizes, both common in the real world. We began by testing at Layer 4, then tested atLayer 7, where – based on previous ZXTM tests – we would expect to see someperformance loss compared with the Layer 4 figures but very little as a percentage. WithMS Virtual Server we achieved a maximum of 5660rps at Layer 4 and 4110rps at Layer 7(http requests).L4/7 Requests Per Second6000566050004110358040002850MS VS L43000MS VS L717302000148010000Requests/second2k file requests/second8k file requests/secondFigure 4 – RPS At Layer 4/7 Broadband-Testing 1995-2007 5

Zeus ZXTM Virtual Mode – Microsoft Virtual Server ReportTest: SSL PerformanceAs more and more secure services appear on the Internet, so more and more SSLtransactions will occur. Real-world SSL transaction performance, whether terminating atthe device under test, or then re-encrypting to send the request onto a target server, istherefore increasingly vital. So it is important for a Layer 7 device to offer a high level ofsustainable SSL support – where the device terminates the SSL sessions, rather thanpassing them to the target server. SSL session handling is an absolute killer for serverperformance, so this is a vital function of a Layer 7 traffic management device.For the SSL tests, we ran two separate tests – using the RC4-MD5 cipher - with twodifferent session ID reuse values – this is the number of times an SSL connection isallowed to use the same ID. For reasons of working the test device as hard as possible,our first test allowed a once-only use of an ID. This is the most secure methodology but israrely used in the real world. It is an excellent way of measuring the capabilities of aLayer 7 device under the most trying circumstances however. We also ran a – moretypical – 10-session ID reuse test in order to get something much closer to real worldfigures. Below we see the Connections Per Second figure for the Ms Virtual Serverdeployment reach 397cps. Note that it’s not long ago since dedicated hardware platformsexceeding 800cps SSL performance were deemed as exceptional, so in real world termsthis a very usable performance figure. With session ID re-use set to 10, we see asignificant improvement in SSL transaction performance, with MS Virtual Server max’ingat 3354cps.SSL CPSSID Reuse 0SID Reuse 10335435003000Conns Per Sec2435250020001330150010003975000SSL CPS MaxSSL CPS 2kSSL CPS 8kFigure 5 – SSL Conns Per Second – ID Reuse 0/106 Broadband-Testing 1995-2007

Zeus ZXTM Virtual Mode – Microsoft Virtual Server ReportTest: Cache PerformanceHere we tested the cache performance of ZXTM, both in terms of Requests Per Second(RPS) achieved and throughput generated with the MS Virtual Server deploymentachieving a max of 8470rps while sustaining a throughput rate of 269Mbps.Cache 000.269Cacherequests/secondCache throughput(GBits/sec)Figure 6 – Cache Performance Broadband-Testing 1995-2007 7

Zeus ZXTM Virtual Mode – Microsoft Virtual Server ReportSUMMARY AND CONCLUSIONSWithin the scope of this test we set out to show, primarily, that the ZXTM trafficmanagement application is able to run, and run well, in a virtual environment.Our testing proved that this is indeed the case, subjecting – as we did – ZXTM to lengthyand repeated tests in a Microsoft Virtual Server environment. What this means is thatLayer 7 traffic management solutions can be deployed in an incredibly cost-effective andefficient manner by those users valuing the intelligence and manageability of such adeployment over out and out performance. The Microsoft solution was stable – we ranextended tests to prove this. However, the forthcoming releases of MS Virtual Servertechnology (it will support 64-bit guests and multiple-CPU configurations) should helpincrease performance significantly.In A Nutshell We proved that it is entirely possible to run ZXTM in a virtualised environment alongsideother applications on the same server. So where out and out performance is less of anissue than data management and manipulation, ZXTM can be a very cost-effective add-onto an existing application server environment.8 Broadband-Testing 1995-2007

Zeus ZXTM Virtual Mode – Microsoft Virtual Server ReportAPPENDIX: MORE TESTBED DETAILSInternet architectures are becoming increasingly complex.Whether you're building network equipment or providing a service, you must deliverconsistent performance under all conditions. Until now, capacity assessment at high-loadshas been a costly and complex process. For this reason Spirent Communicationsintroduced the Avalanche and Reflector appliances to assist with the challenge. AtBroadband-Testing we have taken these web application simulation and planning productsand integrated them into our test-bed simulating real-life Internet conditions; those thatthe average user experiences daily.Figure 7 – Spirent Avalanche 2500Avalanche is described by Spirent as a capacity assessment product that challenges anycomputing infrastructure or network device to stand up to the real-world load andcomplexity of the Internet or intranets The system determines the architecturaleffectiveness, points of failure, and the performance capabilities of a network or system.Using Avalanche to generate Internet user traffic and Reflector to emulate large clustersof data servers, you can simulate even the world's largest customer environments. Thesystem provides invaluable information about a site's architectural effectiveness, points offailure, modes of performance degradation, robustness under critical load, and potentialperformance bottlenecks. It is able to set up, transfer data over, and tear downconnections at very high rates - all while handling cookies, IP masquerading for largenumbers of addresses, and traversing tens of thousands of URLs.Avalanche initiates and maintains more than a million concurrent connections, eachappearing to come from a different IP address. This allows realistic and accurate capacityassessment of routers, firewalls, load-balancing switches, and Web, application, anddatabase servers. It helps identify potential bottlenecks from the router connection all theway to the database. This accuracy is especially critical for gauging Layer 4-7performance. The ability to additionally simulate error conditions such as HTTP aborts,packet loss, and TCP/IP stack idiosyncrasies can help anticipate-and avoid-significant andpreviously unknown impacts on performance. Broadband-Testing 1995-2007 9

Zeus ZXTM Virtual Mode – Microsoft Virtual Server ReportTo enable more accurate load simulations across multi-tiered Web site architectures, thesystem also supports extremely realistic user modelling behaviours such as think times,click stream, and HTTP aborts that cause Web servers to terminate connections whileback-end application servers continue to process requests. Configuring in this way issimple as both Avalanche and Reflector directly from a desktop browser to set up tests,review feedback in real time, and easily reconfigure test parameters.Figure 8 – Creating A Spirent Avalanche TestThe Avalanche also supports browser cookies, html forms, HTTP posts, and SSLencrypted traffic. The system therefore gives you the flexibility to specify data sourcesand mix and match data sets to recreate accurate user behaviour at very highperformance levels.It also simulates SSL loads that can stress the world's most sophisticated secure ecommerce platforms. It also includes configurable cipher suites that enable you toemulate different types of browsers. Avalanche includes a high-accuracy delay factor thatmimics latencies in users' connections by simulating the long-lived connections that tie upnetworking resources. Long-lived, slow links can have a far more detrimental effect onperformance than a large number of short-lived connections, so this approach deliversmore realistic test results.While Avalanche focuses on the client activity, Reflector realistically simulates thebehaviour of large Web, application, and data server environments. Combined withAvalanche it therefore provides a total solution for recreating the world's largest serverenvironments.10 Broadband-Testing 1995-2007

Zeus ZXTM Virtual Mode – Microsoft Virtual Server ReportBy generating accurate and consistent HTTP responses to Avalanche's high volume ofrealistic Internet user requests, Reflector tests to capacity any equipment or network youconnect between the two systems. Its protocol-level accuracy helps you assure thestability and performance of switches, routers, load balancers, firewalls, caches, andother Layer 4-7 devices. The system is ideal for helping infrastructure service providersvalidate, enforce, and maintain service level agreements (SLAs). Broadband-Testing 1995-2007 11

ZXTM from Zeus Technology operates at both Layer 4 (L4) load-balancing and Layer 7 (L7) intelligent traffic management levels and it is Ethernet-based but it is not a switch, or really any kind of "Ethernet device" per se, but effectively a server-based network