HANA is a ‘platform for real-time analysis and application drive.’ It is a next-generation solution realizing real-time business which all users in an enterprise can check and analyze high-capacity data including detailed information with business insight they want at a moment when business transaction occurs.
|Major Functions of Service||Comparison with an Existing DB||Case of Service Use|
|Service Rate||Relevant Additional Information|
1. Structure of no disk input/output
Generally speaking, the most important element for performance of database is to reduce frequency of disk input/output, which is a critical part of database tuning. HANA saves the whole data by making the data as database format in main memory and it has a structure in which disk input/output does not occur at all because data processing and checking is performed within the main memory. This inmemory-based computing can significantly improve performance of database and also enhance parallel processing by random access unlike disk. Just in case of system error, it transmits completed contents of transaction to log volume and changed data on log volume are reflected to data volume by a certain check point. As these disk-related operations including log and data volume are of asynchronous execution from transaction performed in an actual memory, its influence on the performance is small.
2. Column-based saving structure
HANA provides saving structure not only record-based one but also column-based one. According to a report analyzing mass data, mostly only a few parts of columns composing table are used. Due to these characteristics, existing databases only providing record-based saving structure are subject to searching columns unnecessary for operations and cause considerable load on system resources. In contrast, HANA processes operations by using minimum system resources with physical structure which is able to selectively search target columns for analysis and more users can simultaneously use analysis reports on performance improvement and analysis system. SAP has already applied this inmemory-based column saving structure and processing function to BWA (Business Warehouse Accelerator) products and its performance and stability is verified by various examples.
3. Reduction of data saving costs
Efficient compressing technology is required for analyzing the whole data by loading them on memory which is of high cost compared to disk. HANA can considerably reduce space for data saving with characteristics of column-based saving structure and various compression algorithms. The column-based saving structure of HANA first saves columns of identical data properties in consecutive memory space and second executes more efficient compression once more through five compression algorithms. Advantages of this data compression are that performance can be improved by transmitting data with being compressed to operational units and processing them as well as that space for data saving can be reduced. Other than this kind of physical data compression technologies, HANA can reduce overall costs of data management by minimizing total and summary tables which are created for performance improvement by existing databases.
4. Additional analysis function
Although HANA is already capable of performing innovation on information analysis system only with saving structure of inmemory database, it provides additional function needed for analysis tasks. First, modeling with structure of Analytic View and Calculation View is provided for optimized multi-dimensional analysis and analysis function can be improved by using Calculation Engine. Second, if provides Library of various analysis functions, which supports algorithms for basic statistical analysis such as regression analysis, K-means, etc.
Difference between HANA DB and traditional, disk-based DB
|Classification||Traditional approach||SAP HANA DB|
|Data capacity||Column, Saving, Compression (disk-based) Data duplication management (Possessing data, cache and data compression to secure performance on the basis of high-capacity detailed data)||No duplication on data of saving by columns and compressing method (memory-based): Separate summary data mart to additionally secure performance not required. Real-time counting and processing of the necessary business views on memory|
|Speed of reflecting information||Data delay to information system by arrangement operations occurred due to use of ETL tools for extracting various source system data Composing additional summary data mart based on the first loaded high-capacity detailed data by ETL tools, Additional delay in cache operations occurred||Real-time data duplication from various sources through SAP real-time data duplication technology within the unit of 1 second SAP Based on high-capacity detailed data duplicated by RS technology, additional composition of summary data mart not required RS Real-time provision of various business views not by cache technology but by inmemory data processing engine|
|Speed of calculation||Processing with saving method by columns and cache technology, a traditional snapshot of data||Method of compression calculating on all the actual data in memory with data saving method by columns|
|Flexibility||Providing limited flexibility (Fragile to change in data model for saving actual summary data in frame physically composed in advance)||In SAP HANA, actual summary data not saved in physical business view defined in advance and data gathered and dispersed in real time on memory according to virtual business view|
|Application platform||Available to use only for analytical purposes (Unavailable for transaction ones)||Integrating period and information systems|
: Telecommunication company in USA
T Mobile planned aggressive customized marketing by providing appropriate campaigns and plans for major target customers by customer base but there was no technological platform which could promptly collect and analyze monitoring results of major campaigns or plans. In this circumstance, they decided to introduce HANA and automatically collected various customer data created in real time with HANA platform. By analyzing demand and reactions on various campaigns and plans in progress from diverse angles, they could make operation platform for timely and accurate marketing depending on acceptance of new products and campaigns to markets.
: Energy company in UK
Since centrica spent too much time saving and processing high-capacity data in real time with a circumstance of adopting Smart Grid, they were trying to find solutions to increase energy usage efficiency by monitoring customers’ energy usage efficiency by time or to reduce production costs. By introducing HANA at the moment, they could build platform to collect and divide Smart Grid data at 30-minute intervals and immediately check major energy indicators with processing data of huge scale caused in real time. In addition, they could increase energy consumption efficiency by regional groups and provide customized plans by customers e.g. households and businesses. With exact processing on high-capacity data in real time, it is evaluated as a good example that they innovate their business with more than 500 times as improved operations on statistical analysis as before.
: Medical device company in USA
Medtronic’s customer data was dispersed in too many systems including SAP system and Non-SAP systems and there was no analytical system for prompt analysis on whole data even if they integrated all of them in one repository. They were challenged to make and perform optimized customer strategies by customers when needed. Medtronic selected HANA for an integrated analytical platform for integrated analysis on data sources of SAP and Non-SAP systems and made an environment for integrated analysis. They standardized sales reporting process by analyzing this high-capacity data in real time and they had an opportunity to dramatically increase their sales and business profits by grasping and responding to customers’ needs in advance.
※ Rate of service is subject to slight change because it applies exchange rate at the moment of billing.
Slave 1, 2
|SW||OS: SuSE Ent. Linux 11
DB: SAP HANA DB
|Classification||Monthly rate||Hourly rate||Note|
|2 * 20||90GB||$202||$0.281|
|4 * 32||140GB||$404||$0.561|
|8 * 64||230GB||$808||$1.121|
|OS||Monthly Rate||Hourly Rate|
|SuSE Enterprise Linux 11||$118||$0.164|
Network plans (same as server)
|Monthly usage less than 100 GB per Cloud server||Free|
|Monthly usage by customer (on an ID basis) over 100 GB||Refer to the table below.|
|Classification||Usage Section||Application by Cloud server||Application by Customer (by ID)|
|per Cloud server Less than 1 TB||1 TB ~ 10 TB||10 TB ~ 20 TB||Over 20 TB|
|Meter Rate||Unit cost by section(per GB)||Free||$0.082||$0.073||$0.064|
|Flat Rate(by month)||30 TB/month||$1,802|
|50 TB/ month||$2,703|
Add Disk and Public IP
|Additional official IP||Per 1 IP||$5|