100 likes | 374 Views
Middleware:. Gluing the Warehouse Together. Chapter 19 by Gausden and Mason from Data Warehouse , Bischoff & Alexander. Business Issues for Middleware. Definition: software that shields users and developers from differences in services and resources used by applications
E N D
Middleware: Gluing the Warehouse Together Chapter 19 by Gausden and Mason from Data Warehouse, Bischoff & Alexander
Business Issues for Middleware • Definition: software that shields users and developers from differences in services and resources used by applications • Data warehouses often have heterogeneous databases, operating systems, networks, hardware, applications
Business Issues - 2 • Role of middleware • Assist developer in data extraction/transformation and populating DW • Assist business user in accessing DW • Therefore needed at different points in life cycle • Types • Copy management: data extraction, transformation, replication, and propagation • Gateways: DB and independent gateways • Program-to program: RPCs, TP monitors, ORBs • Message-oriented
Populating the Data Warehouse • Connect tool to data (networking or communications protocol) • Access to databases (access method for connection and update) • Data sources • Populating DW databases • Maintenance (replication)
Connectivity and Interoperability • Communications gateways • Protocols – e.g., TCP/IP, DECnet, NetBIOS, ODBC, SPC/IX, Async, OBDC, DRDA (LU0, LU2, LU6.2) • NOSs – e.g., SNA, Windows, OS/2, UNIX, MVS, VMS, Netware, LAN Server, Banyan • DBs – SQL dialects • Feasibility • Multiplatform, multivendor • Versions, upgrades • Solutions • Single vendor • Multivendor • DBMS-independent vendors • Architecture standards, technical skills
Proof of Concept • Start with a pilot project • Staffing • Vendor support • Installation verification • Define business problem; verify scope • Verify technical feasibility & requisite products • Install & customize components on source(s) • Install & customize components on target(s) • Set up security • Install & customize components on clients • Define data sources and targets to global catalog • Identify/set up predefined joins, views, rpcs • Test access to source, creation of targets, access to DW
Front End Tools • Production queries • Access for existing tools • Ad hoc queries • “Intelligent” global optimization • Query governor – preset limit • Predictive governor – estimates cost (CPU, I/O) • Tool connectivity to all databases
Accessing DW Databases • Heterogeneous DBs, linking data marts • Gateways • Database gateway (requires DBMS) • Independent gateway • Aspects • Point-to-point, point-to-many-points • Data location transparency • Global metadata catalog • Access to distributed databases • Heterogeneous joins • Global optimizer • SMP
Planning for the Future • Scalability • New platform? SMP? • Maintenance • Updates, fixes, stability of product line • Changing business environment • Mergers, re-orgs, consolidating data marts
Summary • Heterogeneous software & hardware environment • Heterogeneous • Data sources • Data targets • Platforms • Operating systems • Communications protocols • Connectivity