Successful use of large data is the key to business change

Large data era has come, how to deal with and use of huge information, many enterprises are facing new problems.

Nowadays, a lot of social activities and enterprises are inseparable from it, and in these activities it is necessary to produce huge information. With the increasing popularity of mobile broadband and mobile communication products, this trend has been accelerated. "Information explosion" requires the enterprise system to correctly analyze and handle a large number of complex data, but it is difficult to rely on the previous technology.

Only enterprises that can solve big data problems can get business opportunities from the commercial change.

You know, the variety of big data, for example, in the financial transactions, purchase, service activities, in the recorder and product manufacturing process will produce data, if the rapid processing of these data, not only the existing products and services to improve efficiency, and the past, the artificial business also through computer automation, To improve or provide new services. As a result, new opportunities will be found in different business areas.

Systems that handle large data are:

Only when the above 4 conditions are met can the enterprise obtain commercial benefits from the large data. Previous technologies have been able to handle hundreds of GB of data, but when it comes to processing millions of trillion-scale data, there are problems such as inability to cope, inadequate processing capacity, inability to scale flexibly, and cost.

From the technical level, in the premise of satisfying commercial interests, the key to solve the above problem is that memory takes the method of processing data using RAM (random access storage). Existing performance requirements cannot be met on previous hard drives. Even faster solid-state drives (SSD) and flash memory are far from enough. Can take advantage of RAM that is more cost-effective than HDD. RAM has unmatched performance with HDD, SSD, and flash memory.

Also, software Java is the key to solving many problems. Java has the fine data management function of garbage collection (GC), that is, when the memory space that needs to be allocated is no longer in use, the JVM invokes the garbage collection mechanism to reclaim the memory space, thereby greatly increasing the efficiency of memory utilization. But the GC also has its problem, that is, the application software stops when it is released through GC memory.

Although with the popularity of the "bit" CPU, a considerable amount of memory can be allocated, but the larger the amount of storage, due to the GC caused by the longer the software stop time. This is a fatal weakness for systems that require higher performance. A system that handles large data must solve this problem.

Terracotta has successfully developed technologies that can quickly handle large data. While leveraging Java applications, more than 500,000 of business users, including Fortune 2000, have adopted large data solutions and achieved significant results.

Terracotta offers a solution with two core products. One is the Enterprise Ehcache (cache), which is used in the Java Business System to store large amounts of data in an external cache memory to improve the performance of the application software. The second is bigmemory, which supports a process-by-heap cache that can be used to store application-related bulk data.

The Memory data management environment implemented by these two products does not rely on specific operating systems and Java environments and databases. When the two products are installed between the software server and the data storage area, the various database databases in the rear become an abstract and huge database, which can quickly process the data and files in the database.

When installed in the Java memory domain, the preceding question of GC (garbage collection mechanism, automatic memory) is solved.

At the same time, this data management environment is run on the product server, and the data is placed in the area scale up by adding servers. Through this method, the above 4 problems can be solved at the same time.

Terracotta customers, such as Visa and PayPal, which provides online payment services, use terracotta's large data solution in order to solve the problem of misconduct caused by credit card and online billing.

It is necessary to analyze the large data at high speed in the process of checking the behavior. In the past, although the use of expensive hardware and database, but its effect is unsatisfactory. Companies have successfully created an environment for the timely analysis of terabytes of data after using terracotta large data solutions.

Visa used to deal with 1 of irregularities, usually taking more than 10 minutes, and the time taken to take advantage of our large data solution has been shortened to just a few seconds.

In addition, Australia's largest communications company, Telstra, uses CRM (Customer Relationship Management system) to manage the vast majority of customer information. When using large computers and open environment RDB (relational databases) to process CRM, it takes an average of 30 seconds to process a piece of information because of the need to refer to the customer in the database. Because 95% of the clicks between the company and the customer come from the Customer service center, the company invests 1 billion dollars a year in the operation of the Customer service center. The company has considered to transfer the customer service center to the website, but it will take 30 seconds for reference to customer information to give up.

In this case, the company started the terracotta large data solution, took the database of customer data memory before processing. In this way, the time required for reference processing is successfully shortened to 30 milliseconds. It is expected that the size of the company's customer information will increase from 10TB to 100TB in a few years ' time.

In this way, when the speed of the TELSTRACRM system is improved, the service is improved while the cost is successfully reduced. The Company believes that the use of terracotta large data solutions will solve the problem of cancellation, and through the implementation of cross-selling, a substantial increase in sales.

This shows that the company's memory database management solutions, not only can solve the past problems, but also can be a great impetus to promote business change.

The writer is the general manager of Terracotta, an expert in business software, with more than 20 years of working experience. Under his leadership, terracotta has been remarkably successful in dealing with large data. At present, terracotta for many of the world because of "information explosion" and facing business change enterprises to provide solutions.

(Responsible editor: Lu Guang)