The realization of cloudification of IT architecture has become the general trend of enterprise IT strategy. Whether private cloud technology or public cloud technology is adopted, the software is required to have the adaptability of the cloud environment. As the most important data asset of an enterprise, it relies on the underlying data management software for effective management. In order to obtain practical knowledge and information from massive data, the technology of how to efficiently organize the storage and search of data has been evolving. From the early hierarchical database to relational database, from SQL database to NoSQL database to Hadoop , graph database and other platforms for processing unstructured data, it has been developing and changing. Now it can be said that a hundred flowers are blooming and a hundred schools of thought contend.
With the rise of cloud technology, there are more choices for the location of data storage. Data may be stored in the user's data center or in a public cloud provider platform. Therefore, for users, how to maintain the consistency of experience in different infrastructures, different database storage platforms, and on and off the cloud, smooth application access, development, migration and data security, and protect investment in IT skills , becomes extremely important. IBM's next-generation data analytics platform has been redesigned and developed to achieve the following blueprint.
For users, information visualization tools may use popular products such as Cognos or Tableau, data mining may use tools such as SPSS or R, and development languages may use popular languages such as Python or Scala. Data may be stored in traditional relational databases such as DB2, Oracle or PDA, or in newer Hbase, Hive or NoSQL databases. Emerging Internet customer data may be naturally stored in cloud databases and so on. Can we design a general engine, which can not only provide a unified interface, shield the underlying data storage heterogeneity, but also use a unified SQL-based language for application access and development? The answer is yes! Based on the open source Spark technology, IBM has developed the next-generation general-purpose database engine Common SQL Analytics Engine and the query access engine Fluid Query Engine.
As an early DB2 relational database, in order to flexibly support zOS, Unix/Risc architecture or WinTel architecture, and at the same time to meet customers' elastic expansion and support different workloads, the concept of Universal Database has been proposed since version 7, that is, a unified The database SQL engine can support different architectures. After versions 8, 9, and 10, this architecture has been greatly expanded. Through the expansion of functional modules such as HADR, DPF, PureScale, BLU, and GDPC, under the unified SQL engine, it perfectly supports the differences of users. workload.
However, the IT architecture is constantly evolving, and the cloud environment has also developed architectures such as virtualization technology and container technology to support massive access, multi-tenant access, and rapid deployment expansion. The development of the mobile Internet has greatly promoted the interaction with users to generate a large amount of unstructured social media, voice, image, video or sensor data, so there are more diverse forms of data. In order to solve the above challenges, IBM has also provided solutions such as BigInsights, a NoSQL database or an all-in-one data warehouse PDA based on open source technologies, through technological innovation or acquisitions. However, with the accelerated pace of customer cloudification, the architectural differences between the above solutions have gradually become a technical bottleneck for users. The implementation of data management in each scheme is different, which leads to the increasing difficulty of DBA management, operation and maintenance and application development. Based on the needs of cloud-based applications such as the mobile Internet, a more flexible architecture is required to support the challenges of massive access, sudden user growth, 7*24 uninterrupted access, and agile development. For this purpose, IBM has adopted the strategy of general analysis engine in the design of the next generation analysis engine, that is, unified analysis engine, to support users to deploy data applications in different environments, regardless of whether they will be deployed in the architecture of the original data center, Or a private cloud or even a public cloud environment. This kind of organization will provide users with great flexibility. The code can be written once and can be run in any environment, while ensuring the consistent operation and maintenance experience of the DBA!
The first goal is to unify IBM's own data repository code. The newly released cloud relational database engine dashDB, the classic DB2 11 version, the dashDB on Local suitable for private cloud deployment, IBM Hadoop BigInsights and the upcoming next-generation data warehouse appliance PDA will all be built on unified analytics. On the engine code, users will be able to write code only once and deploy and run anywhere. For DBAs, no matter which library is operated in, the operation processing mode is unified. For users already deployed on a certain platform, data and applications will migrate smoothly in the future, whether they migrate to the cloud or build their own private cloud environment. For tool developers, only one platform is certified, and all other platforms can be smoothly supported.
Therefore, whether users are still using the traditional Power architecture and adopting DB2 to support their core trading system, they can gradually upgrade to the DB2 11 version to enjoy the general analysis engine capability of transitioning to the cloud in the future. If users are about to deploy a private cloud environment, they can use the dashDB on Local solution. It is deployed based on containers and can support massive expansion of petabyte-level data. It also has general SQL database engine capabilities, operation and maintenance experience, development interfaces, and its core transaction DB2. be consistent. Or directly adopt the PDA solution, which integrates software and hardware implementation, which can simplify operation and maintenance management, but the management experience and development method are also consistent with the database of the core system. If users develop various applications directly based on public cloud services, the dashDB cloud data service on it is consistent with the database management and development experience under the cloud. Finally, even if the user chooses to use Hadoop to store their unstructured data, the built-in BigSQL engine, its operation and maintenance management, and the development interface maintain the same experience as the core transaction database they are using! See now, do you feel very Cool! Whether it is transaction load or analysis load, whether it is used under the cloud or development and operation and maintenance on the cloud, all skills and software assets are protected, and your cloudification plan is fully guaranteed!
For data that has been stored in different databases, it is not realistic to formulate a unified data storage strategy. Therefore, data storage is discrete, and multi-database storage will be the most likely reality. For users, how to implement a unified access interface, transparently access all data sources, or migrate data between different data sources will become a key challenge. IBM provides Fluid Query Engine, a general engine for data access, which fully considers the reality of customer data virtualization, unifies interfaces, and solves customer problems.
As a result, through Common Analytics Engine and Fluid Query Engine, the implementation code of the infrastructure is unified, and the interface for applications to access different data is simplified, so as to achieve a consistent user experience of data access under the cloud and on the cloud, protect existing development IT assets, and help users to early Realize the cloud plan!
Come experience it, download and try IBM Analytics Platform products!
Deadline: October 30, 2016
For details, please consult online customer service !
Customer Service Hotline: 023-66090381