Oracle Database In-Memory: breaks through the row and column restrictions to achieve real-time analysis

Source: Internet
Author: User

Oracle Database In-Memory: breaks through the row and column restrictions to achieve real-time analysis

Two weeks ago, Oracle released Oracle Database In-Memory globally. Oracle Database In-Memory will be provided as an optional accessory for Oracle databases, it can run on the system where the enterprise runs the Oracle database. It supports applications currently in use by the enterprise without any modifications to the application.

Andrew Mendelsohn, executive vice president of Oracle Database technology products, also came to Beijing to explain Oracle Database In-Memory to Chinese media. According to reports, Oracle Database In-Memory has three major objectives: real-time analysis and performance improvement of 100 times; faster OLTP hybrid workloads; and easy migration, easy to deploy without changing the application. Andrew Mendelsoh emphasized that improving the analysis speed is the most important goal.

Andrew Mendelsohn, executive vice president of Oracle Database Technology Products

Database memory in dual format of rows and columns

In the past, customers had to select only one type between a row-based database and a column-based database, thus being limited in data processing performance. Row-based databases are more suitable for processing transactional data, such as applications in e-commerce, inserting or querying a sales order, while column-based databases are better at analysis operations, such as generating sales reports by region. One of the major innovations of Oracle Database In-Memory is to break through the row and column restrictions. It uses dual-format Database Memory and supports both row and column formats In the Memory of the same table, activate and maintain transaction consistency at the same time. The analysis and reports are in the column format, while the OLTP is in the row format.

In the Oracle memory columnar storage technology, pure in-memory columnar storage can quickly respond to data changes and achieve a compression ratio of 2 to 20 times, it also supports table-level and partition-level granularity, and is suitable for all mainstream hardware platforms, facilitating enterprise deployment. Andrew Mendelsohn stressed that "SAP needs to purchase hardware to support the memory columnar storage technology. For example, if customers want to purchase sap hana, But Oracle supports all mainstream hardware platforms, customers do not need to purchase hardware products."

During the test, the per-CPU kernel of the column format can scan 1 billion entries per second, while the row format can only reach 1 million entries, and the performance can be improved by more than one hundred times. In addition, the table connection speed is also increased by 10 times by converting multi-table join operations to efficient column scanning.

For traditional OLTP systems, analyticdb indexes are often used for fast query. In this architecture, inserting a record into a table requires updating dozens of indexes at the same time, OLTP system performance is forced to decrease. By replacing analyticdb indexes with column store, Oracle can quickly analyze any column in the New OLTP system, improving the OLTP and batch processing speeds.

Software and Hardware Integration

Oracle Database In-Memory can meet any needs through vertical scaling on large SMP servers, horizontal scaling across Server clusters, and storage layering. It can run databases of any scale economically and cost-effectively, whether it is a PB-level data warehouse, big data processing, or database cloud. Oracle Database In-Memory also supports Java, R, big data, PHP Python, Node, REST, Ruby, and other languages. It fully supports various data types, including relational, object, XML, text, spatial data, and the latest integrated JSON.

The Oracle integration system perfectly complements Oracle Database In-Memory. Oracle integration systems, including Oracle Exadata Database ECs and Oracle SuperCluster, are optimized for Oracle Database In-Memory to provide large-capacity storage, extreme performance, and high availability, it also provides storage layering to transfer less active data to flash and disks to provide outstanding cost effectiveness. The memory Fault Tolerance function of the Oracle Integration System selectively copies memory data across multiple nodes. Therefore, if a server fails, the memory data copy can be used immediately to meet the query requirements. The latest Direct-to-Wire Infiniband improves the Memory Scale-out speed.

Oracle M6-32 Big Memory Machine Oracle M6-32 Big Memory Machine is a vertically scalable platform for Oracle Database In-Memory, providing up to 32 tb dram Memory and 3 TB/second Memory bandwidth, improves memory performance to the maximum extent.

In addition, Oracle Database In-Memory can be deployed easily and quickly In any existing application environment compatible with Oracle databases, you do not need to make any changes to the application. Andrew Mendelsoh said that it is easy to deploy Oracle Database In-Memory. The Memory computing function can be enabled through the "inmemory_size = xxx gb" command statement. Mark Rittman Mead, chief technology officer, said: "There is nothing to say about the use of Oracle Database In-Memory. Just turn on the switch and load the table to the Memory, nothing else needs to be done."

Andrew Mendelsohn said: "As the only vendor in the industry that can fully implement memory computing across applications, middleware, databases and systems, Oracle is committed to promoting memory technology to help enterprises improve database performance and make informed decisions, and promote sustainable revenue. The advent of Oracle Database In-Memory is the result of Oracle's continuous innovation and investment In the Database field, and also the fulfillment of Oracle's commitment to enterprise customers around the world. We hope to provide the ultimate availability, scalability and security for all customers at the lowest deployment and O & M costs ."

Related Article

Contact Us

The content source of this page is from Internet, which doesn't represent Alibaba Cloud's opinion; products and services mentioned on that page don't have any relationship with Alibaba Cloud. If the content of the page makes you feel confusing, please write us an email, we will handle the problem within 5 days after receiving your email.

If you find any instances of plagiarism from the community, please send an email to: info-contact@alibabacloud.com and provide relevant evidence. A staff member will contact you within 5 working days.

A Free Trial That Lets You Build Big!

Start building with 50+ products and up to 12 months usage for Elastic Compute Service

  • Sales Support

    1 on 1 presale consultation

  • After-Sales Support

    24/7 Technical Support 6 Free Tickets per Quarter Faster Response

  • Alibaba Cloud offers highly flexible support services tailored to meet your exact needs.