Presentation on theme: "Evaluation of performance aspects of the Auto-ID Infrastructure"— Presentation transcript:
1 Evaluation of performance aspects of the Auto-ID Infrastructure Kai Sachs (TU Darmstadt) Supervisors: Christof Bornhoevd (SAP)Mariano Cilia (TU Darmstadt)Evaluation of performance aspects of the Auto-ID Infrastructure
2 CONTENTS Auto-ID Infrastructure Measurement Approach Results of the ExperimentsFinal Conclusions
3 Auto-ID Infrastructure Measurement ApproachResults of the ExperimentsFinal Conclusions
4 SAP Auto-ID Infrastructure 2.0 (AII) AII: Overview (1)SAP Auto-ID Infrastructure 2.0 (AII)Middleware solutionReceiving RFID data from data capture sources (e.g. RFID devices)Integrates the data into enterprise applications.Early prototype
5 SAP Auto-ID Infrastructure (AII) Auto-ID Cockpit (Web User Interface) AII: Overview (2)The illustration below shows an overview of SAP RFID landscape:ReaderDeviceControllerSAP Auto-ID Infrastructure (AII)SAP ExchangeInfrastructure (XI)SAP R/3RFIDTagsBackendAIILLIXML/PMLXMLIDocAuto-ID Cockpit (Web User Interface)Traffic GeneratorTraffic GeneratorFrom: SAP RFID Solution Package SAP Auto-ID Infrastructure 2.0 (AII) Theory
6 Integration Layer (XI) Auto-ID Node System ArchitectureAuto-ID CockpitAuto-ID NodeDCBEIDocXMLMessageDispatcherActivitiesXMLIntegration Layer (XI)Communication LayerCommunication LayerXMLTGIDocBERule EngineAINRepositoryFrom: SAP Auto-ID Infrastructure
7 CONTENTS Auto-ID Infrastructure Measurement Approach Results of the ExperimentsFinal Conclusions
9 What should be observed? Experiments settingsMultiple readersMessage sizeSystem behaviorCPU loadIO ActivitiesSingle processesMemory …ThroughputComponents on the Auto – ID InfrastructureGross TimesGross CPU TimesCustomized Traffic GeneratorMicrosoft PerformanceCustomized Traffic GeneratorJARM
10 Microsoft Performance Part of Microsoft Windows 2000 & XPSystem MonitorAllows to observe:Single processesIO ActivitiesCPU load…Observations could be logged in a CSV - file.
11 JARM Allows observation of Java components Provides averages values and sums per componentHierarchies of components are possibleResults are accessible through Visual AdministratorNeeds source code modifications!Problems, if JMS is used
14 Customized Traffic Generator Based on SAP Traffic GeneratorUsed to simulate reader observationsNew logging functions were added Every sent request can be logged Allows better review of throughputOther new functions:Add Timeframes for experimentsSend a defined number of messagesPossibility to run different scripts parallelScenario – Definitions…
15 CONTENTS Auto-ID Infrastructure Measurement approach Results of the ExperimentsConclusion
16 Results of Experiments CPU LoadIO ActivitiesThroughputJ2EE Components of the Auto-ID NodeDifferent VM settingsSettings of Message Dispatcher
17 Results of Experiments CPU LoadIO ActivitiesThroughputJ2EE Components of the Auto-ID NodeDifferent VM settingsSettings of Message Dispatcher
19 CPU LoadIncursions and the observed fall down have heavy influence on the average CPU loadCPU load differ for the experimentsThroughput depends on CPU loadNeed for a key figure for comparison of the different experiments.
22 IO Activities IIIMaxDB Savepoints have a significant influence on the system behavior.Settings for MaxDB Savepoint intervals can be changed.Influence of Savepoints is bigger, if the files are fragmented.The Savepoints could not explain the CPU load fall down in the end of the experiment time frame!!!
23 Different message sizes ThroughputDifferent message sizes9 EPCs per message45 EPCs per message90 EPCs per message900 EPCs per messageMultiple readers1 simulated reader3 simulated readers5 simulated readers7 simulated readers10 simulated Reader
27 Throughput V Conclusions: Influence of message size: Bigger message size Higher throughput in no. of EPCs per sec.Influence of multiple simulated RFID readers:Throughout increases up to n reader; decreases after thatThroughput decreases over time
32 Auto-ID Node Components IV Conclusions:Gross Times scale linear for different message sizes.The activities are the dominating part of the Auto-ID Node.The activities are dominated by database accesses.
33 CONTENTS Auto-ID Infrastructure Measurement Approach Results of the ExperimentsFinal Conclusions
34 Final Conclusions I CPU Load: CPU load has short incursions Number of simulated readers has no influence on the CPU loadMessage size influences the proportions of the system processes regarding CPU loadCPU load decrease at the end of the experiment time frameIO Activities:MaxDB Savepoints have a significant influence on the system behaviorThroughput:Throughput is higher for larger messagesThroughput decreases over timeThroughput depends on number of readers
35 Final Conclusions II Components of the Auto-ID Node: Auto-ID Node components scale linearRule Activities are the dominating componentPerformance of Activities is dominated by database accessesNumber of simulated readers has significant influence on the Gross TimeSettings of Java Virtual Machine:Heap size is the most important parameter for higher throughputJMS settings of Message Dispatcher:Throughput is lower, if JMS is used.Gross Time is higher, if JMS is used.