蓝图：优化金融服务数据交换

Optimizing Financial ServicesData Exchange A PlatformDIGITAL®Pervasive Datacenter Architecture (PDx®)BLUEPRINT PDx®Blueprint:OptimizingFinancial ServicesData Exchange Introduction Traditional IT architectures withinFinancial Services (FinServ) companiesaren’t designed to effectively leveragedata, optimize data exchange or addressthe challenges of Data Gravity. Datacreated from different lines of businessis often stored in silos throughout thecompany, while some elements areon-premise, and some are in the cloud.This distribution without intent leadsto performance issues, operationaland supply chain costs, and increasedoverall complexity. Successful digitaltransformation requires a data-centric IT infrastructure that localizesdata aggregation, staging, analytics,streaming and management in centersof data exchange at global points ofbusiness presence. Inconsistent data storageand access methodsleading tostorage sprawl, cost overrunsand compliance issues Implement distributed datastaging/aggregationto optimizedata exchange between users, things,networks and clouds Fragmented architecturesareburdened by technical debt and canbe driven by point solutions lackingthe capabilities and performancerequired for hybrid IT workflows Integrate public/private datasourcesto enable real-timeintelligence across distributedworkflows Distribute business intelligencecapabilitiesto allow secureB2B data exchanges and unlocknew opportunities Deploy regional data lakes/distributed data warehousesto maintain data performance,compliance and sovereignty Siloed datapreventing theenablement of analytics andnew business models centeredaround data Cloud connectivity and networkaren’t optimized, causing poorapplication performance whenleveraging the cloud to accesslocal data Solution Step 1 Step 3 Step 2 Implement data staging/aggregation Host data and analytics adjacent to networkingress/egress Integrate public/private data sources 1Distributed data staging/aggregation2Regionalized data storage for compliance1 3Integrated public and private data sources Action:Directly interconnect cloud on-ramps to centers ofdata storage Action:Implement a cohesive data storage strategy at centersof data exchange Step 1:Implement data staging/aggregation 1Deploy centers of datastaging in key locations2Data Lakes store raw datato be analyzed and curatedby data scientists3Refined data sits in the datawarehouse for businessprofessionals to use4Due to the value andsensitivity of enterprisedata, access needs tobe strictly controlledand logged Action Outcomes Step 2:Integrate public/private data sources The Core SwitchingInfrastructure terminatesconnectivity into theData Hub and enablesaccess to the cloud andother data sources bydirect high-performanceinterconnection Integrated public andprivate data sources DATA HUB Action Additional connectivityis provided by use ofsoftware-defined on-rampssuch as ServiceFabric™ Other data sources canbe cloud storage, IaaSenvironments, SaaSenvironments or otherremote Data Hubs Outcome +Optimized data exchange betweenusers, other devices, networksand clouds Step 3:Host data and analytics adjacent to network ingress/egress GPU Farm is locateddirectly adjacent to storeddata for direct accessto enable AI Developmentand workloads Unlocked newbusiness opportunities DATA HUB Action Bulk compute farmis used for media contentcreation, complexmodeling and simulations Distribute business intelligence andconnect global data ecosystems by: +Adding processing, analytics andstreaming capability at global pointsof business presence+Hosting a B2B meeting place forcompanies to collaborate and connecttheir business platforms Outcome +Enabled real-time intelligenceacross distributed workflows locallyand globally Target state architecture Summary A purpose-built data-centric architecture to optimize FinServ dataexchange can reduce security risks, lower costs as a result of areduction of bandwidth and duplicated infrastructure, and contributeto revenue growth through unbounded data analytic performance.This is necessary to support the exploding volume, variability andvelocity of data creation, and the processing and storage requiredto accommodate digital business. This strategy brings the users,networks, systems and controls to the data, which removes barriersof Data Gravity and creates centers of data exchange to scale digitalbusiness. This Optimizing Financial Services Data Exchange Blueprint is part ofa library of blueprints and repeatable implementation patterns thatmake up Digital Realty’s Pervasive Datacenter Architecture (PDx®).Created by practitioners for practitioners, PDx®codifies hundredsof production deployment combinations to enable companiesto accelerate deployment and improve the precision of theirinfrastructure to scale digital business globally.1PDx®provides a step-by-step strategy for FinServ firms as they architect a decentralizeddata-centric IT infrastructure to remove Data Grav