This paper mainly introduces the ISAS5710 system for Data mart and ODS application, and takes ISAS5710 Medium system as an example, it focuses on how to install and configure the ISAS5710 system, how to design and deploy the database of User data mart and relevant analysis and application, To help you quickly learn the basics of using ISAS5710 Rapid Deployment Data mart applications.
With the continuous improvement of the user's business system and the increasingly fierce market competition, more and more enterprises are building data Warehouse, Data mart, ODS system to analyze the user's demand, optimize the operation of the enterprise, find more market opportunities and be in the leading position in the competition. When users implement a data warehouse, Data mart, ODS system, for users, a very big challenge is the future Data Warehouse system should choose what kind of Data warehouse software, choose what kind of server, server configuration, select how many servers, choose what kind of storage device, What is the storage capacity, what is the storage device configuration, and what kind of network device it is configured to ensure efficient and stable system performance. At the same time, with the application of the system, the amount of data will increase dramatically, how to provide better system scalability under the premise of ensuring system performance is also a matter of great concern to users. In addition, for most users, want to be able to provide a pre-configured, performance optimization of the hardware and software integrated machine, to achieve configuration, optimization, service one-stop service.
To address these issues, IBM combines its many years of experience with customer data warehousing systems and, in collaboration with IBM software departments, server departments, storage departments, and laboratories, has launched the IBM Smart Analytics System (ISAS) solution, a server, storage, A complete solution, including data warehousing software, based on IBM best practices and fully validated, is a pre-configured, performance-optimized, instantly available solution that customers do not need to rely on guesswork or illusion to configure and validate, to implement out-of-the-box. IBM Smart Analytics System employs the idea of a balanced configuration BCU (balanced revisit unit) that provides balanced performance for each component (database, server, and storage) to ensure optimal overall solution performance. At the same time, it uses scalable modular design, the Data Warehouse system in the whole life cycle, can be incrementally expanded to achieve predictable and measurable performance.
IBM Smart Analytics System Architecture
IBM Smart Analytics System (ISAS) consists of the following modules:
Figure 1. ISAS Architecture
Foundation Module: Sometimes called administration BCU, it is assigned to a single database partition. The module mainly includes cataloging partition, coordination partition and single partition table. The cataloging partition is used to save the system catalog table, the coordinated partition (coordinate partition) is mainly responsible for processing the user's request, and according to the partition key, the user's request is decomposed into multiple subtasks to be processed in parallel by different partitions, Finally, the execution results of different partitions are returned to the user. The system must have 1 Foundation Module. Data Module: Sometimes called the data BCU, which is typically assigned to multiple database partitions. This module mainly saves the user's production data. The system must have at least 1 Data Module. Depending on the amount of data, we can allocate multiple data Module. User module: It is usually used as an additional DB2 partition coordination node to balance the user's connection, and if the system has a large number of user access, we can consider adding user module. The module is optional. Failover module: It is mainly used to meet the requirements of HA, Failover module is similar to Data module, but it does not contain disk storage. The Failover module has been in a standby state and is used to replace the failed administration module, User module, or data modules. Tivoli? BAE Don for multiplatforms continuously monitors DB2 resource usage, including hardware and software, and uses Failover module to replace failed modules in a timely manner to ensure high availability of the system. Warehouse Applications Module: It is primarily used to run Infosphere Warehouse Data Warehouse software. The Warehouse applications module is compatible with the business FDI module, and is the IBM Smart Analytics System. Provide data analysis capabilities. Warehouse Applications module can contain 1 or 2 nodes, where warehouse applications node is required and warehouse OLAP node is optional. Warehouse appLications node contains all the components in creator server tier in the Infosphere Warehouse software, including Infosphere Warehouse administration Console and SQL warehousing Tool (SQW). If an optional OLAP node does not exist, cubing Services can also run on that node. Warehouse OLAP node is optional, it has two features: Running cubing Services Cube Server in a high OLAP usage scenario, providing active-active high availability, and any node failures in that module can take over each other. Business FDI module: It is primarily used to run the Cognos Business FDI software to provide reporting and data analysis capabilities. Depending on the usage, it can contain 1 or more nodes.
Depending on the size of the user data and the number of user visits, ISAS provides a variety of models to fully meet the needs of user data analysis and achieve optimal processing performance. As shown below:
Table 1. ISAS Solutions
-5710 5600 7700 9600 positioning cost-effective data analysis, report solutions, out-of-the-box, suitable for small and medium market and departmental applications for cost-effective flexible data analysis systems, optional Solid state drive– reduced time delay for scalable to hundreds of TB Level data scale environment, standard Solid state drive– reduced time delay advanced query/Workload management system, optimized for disk controller – reduced time delay Server IBM System x3630 M3 IBM System x3650 M2 IBM Power 740 IBM System Z10 Storage IBM System Storage DS3524 IBM System Storage DS3500 IBM System Storage DS3500 IBM System Storage DS8700 operating system SUSE Linux Enterprise Server 11sp1 SUSE linux Enterprise server SP3 AIX 6.1 z/Os V1.12 software infosphere Warehouse departmental Edition,cognos Business FDI infosphere Warehouse Enterprise edition,cognos Business FDI (Optional) Infosphere Warehouse Enterprise Edition,cognos Business FDI Reporting (Optional) DB2 for z/OS, infosphere Warehouse for System Z,cognos Business FDI Reporting (Optional) analysis capabilities cubing Services, Query and Reporting, Data Mining, text Analytics cubing services, Business FDI, Data Mining, text Analytics cubing Services, Business FDI, Data MiniNg, Text Analytics cubing services, Business FDI linear extensibility No Yes Yes build, deploy, health check and solution support Services YES,IBM Partners can provide build, deploy, health checks Check Service Yes Yes
According to statistics, in the current data warehouse market, nearly 65% of data marts or data Warehouse systems currently have about 3TB or less of data, and to this end, IBM launches ISAS5710 to meet the needs of data analysis for data marts, small and medium sized enterprises, and company-level applications, while providing a higher price/performance ratio.
