What Makes HUAWEI CLOUD DWS the First Choice for Enterprise-Level Data Warehouses?
Dec 01, 2018
DWS is among the first batch of products in China that entered the Forrester Wave report, and has achieved wide recognition in the industry.
Huawei's enterprise-level data warehouse service (DWS) was recently listed in IT consulting firm Forrester's research report, The Forrester Wave™: Cloud Data Warehouse, Q4 2018.
In both 2017 and early 2018, DWS was recognized in the Gartner Magic Quadrant for Data Management Solutions for Analytics report and ranked first among Chinese brands. These achievements represent the high recognition DWS has received from industry authorities.
DWS has been put into commercial use for 300+ key customers in multiple fields, such as finance, IoV, government and enterprise, e-commerce, energy, and telecom. According to these customers, DWS delivers better querying and analysis performance and is several times more cost-effective than other data warehouses like Oracle. In addition, DWS features large-scale expansion and enterprise-level reliability. What advantages have made it the first choice of so many customers?
Distributed Architecture and MPP Engine: Supporting Agile T+0 Business Decision-Making
DWS is based on the shared-nothing architecture, uses a massively parallel processing (MPP) engine, and supports hybrid row-column storage. Cutting-edge technologies involved in DWS include an intelligent cost-based optimization (CBO) model, B-Tree indexing, late materialization, and LLVM. Equipped with these features, DWS delivers excellent performance and has the ability to support enterprise T+0 data analysis and agile business decision-making. Users can experience millisecond-level responses to analysis of trillions of data records without any optimization.
DWS architecture
Strong Data Consistency and High-Availability Design: Meeting Enterprise Core Service Requirements
To allow company leaders, analysts, and business personnel to view the most accurate first-hand reporting data at any time, data warehouses must ensure data consistency, accuracy, and integrity. Services also need to be running reliably 24/7.
• DWS features atomicity, consistency, isolation, and durability (collectively known as "ACID" in the field of distributed transactions). These properties help DWS ensure data consistency, accuracy, and integrity whenever data is added, deleted, or modified.
• DWS is highly reliable. All of its components use active-active or active/standby design and combine with the SQL fault-tolerant feature, ensuring service continuity when software and hardware damages occur.
• As the volume of data is constantly growing, scaling capabilities have become essential for enterprise data warehouses. DWS supports the online scale-out mechanism, which ensures that services are not affected during the scale-out. DWS is also equipped with an active/standby/slave data protection mechanism and provides an automatic incremental backup function. The backup data is stored on HUAWEI CLOUD OBS with 11 nines of reliability, ensuing zero data loss.
Seamless Integration with Data Lakes for Mining EB-Level Data Values
Enterprises generally have various service systems, O&M systems, and IoT devices. Each system or device generates a large amount of data every day. The data is aggregated to form PB- and EB-level data lakes. Therefore, more and more enterprises are looking to expand their data analysis capabilities from data warehouses to data lakes to perform in-depth analysis and mine more rules and value.
• The Express feature of DWS was developed to meet the requirements. With this feature, users can directly query and analyze the EB-level OBS data lakes without preloading data to DWS. This helps users explore the values of dark data in data lakes.
• The Express feature is fully compatible with the SQL 2003 standard. Existing clients or BI tools can use it without any modification. This feature uses a CBO model and intelligent operator acceleration, enabling highly complicated querying and analysis on EB-level OBS data in minutes or even seconds.
• Users do not need to purchase the Express feature in advance, nor configure or manage it during the use. They only need to pay for the amount of scanned data, which is inexpensive.
Express architecture
Secure and Trustworthy: Resolving Enterprises' Security Concerns
DWS is built based on the software infrastructure of HUAWEI CLOUD, including ECS and OBS, which won the Trusted Cloud Certification issued by the China Data Center Alliance in 2017. In addition, DWS has passed the ICSL certification of Huawei Cyber Security Lab, which is based on security standards issued by UK authorities, and is the highest standard in the industry.
DWS also provides cloud services in Europe and has passed the PSA authentication issued by German authorities, meeting the EU's requirements for data security and privacy.
Next, DWS will release a major security feature, transparent encryption, to provide a more competitive solution for higher user data security.
Ease of Use and "Zero" Learning Costs
Constructing traditional data warehouses takes several days and requires specialized personnel, but with DWS, customers only need to click a few buttons and configure a few settings. The process is hundreds of times more efficient than traditional data warehouse construction.
• DWS provides high, medium, and low specifications. An enterprise that initially selects medium specifications can then increase or decrease the number of nodes or scale the cluster to accommodate a growing number of services. In addition, DWS provides comprehensive O&M management functions, eliminating the need to pay for O&M separately.
• DWS is fully compatible with the SQL 2003 standard, has built-in OLAP analysis functions, and is able to pass TPC-H and TPC-DS standard test sets without any modification. Its wide variety of built-in functions enable complex statistics analysis.
• DWS is compatible with the open-source PostgreSQL ecosystem and seamlessly interconnects with open source IDE, ETL, and BI tools in the industry, such as SQL Workbench/J, DBeaver, Kettle, and Superset. It is also compatible with various mainstream third-party commercial software.