How to use "small data" strategy to deal with big Data challenge?

ManageEngine is a http://www.aliyun.com/zixun/aggregation/13597.html "> Enterprise IT Management software provider, President Raj Sabhlok said:" Nowadays, Large data presents new challenges for the busy IT department, many of which are already more than just a logistical problem in implementing and deploying a new data management platform. The problem with the IT department of the Enterprise has long been transformed from not enough data to data.

"If we go backwards to 10 years ago, from an IT perspective, we have a serious lack of data collection." At that time, we did not have enough data on the systems, networks, and applications associated with the enterprise. "Sabhlok said. "But now, all the data information of the enterprise has been able to be detected." We are in a situation where your enterprise's applications, systems, and devices can alert you to related events and logs. In general, these are the so-called machine data. ”

Is this really a good phenomenon? The more data The enterprise collects, the better? Not necessarily.

"Most of our customers find that collecting so many data doesn't help them very well," he said. , Sabhlok said. "First, they need to spend a lot of effort to maintain this data." So when your business wants to get rid of a bunch of big data, it's a really tough task. ”

In addition to ensuring the availability, performance, and security of enterprise systems and applications, enterprise IT departments often need to support the enterprise's large data strategic tasks. "Companies are trying to use large data analysis to improve and optimize their business, and this analysis is done by sifting through historical data." "Sabhlok said.

Large data can also help companies develop the effectiveness of their internal IT organizations. But often the deluge of data from multiple data sources does not provide a pragmatic, direct benefit to the information. "What businesses need is something that can be used for direct manipulation, and what is needed is to filter all the data sources within the enterprise using a variety of filters, rather than having a comprehensive large data project." "Sabhlok suggests starting with the" small data "attempt, the company really does not need to start a big large data project immediately. Sabhlok told InformationWeek.

Companies need to keep a close eye on the operational data reports of their existing tool functions, which reveals the operational data. "We want companies to take full advantage of the information they already have." "Sabhlok said. "This is what I call the ' small data ' strategy. Start small and do a little bit of work so that you end up with some big results. ”

Security is one area where this pragmatic approach can play its part, says Sabhlok, and requires immediate action by letting it know when a suspicious, security-threatening event occurs. "The IT department needs some kind of advance warning." For example, a sales rep is logged on to a corporate network in New York, but records show that the sales representative logged in at 3 o'clock from a Russian IP address ... "he said."

If an IT enterprise has multiple data sources, but there is no relevant tool to process this data information, thereby helping the enterprise to provide operational insights, this is a potentially huge problem, Sabhlok points out.

The same is true for customers and suppliers. They tend to focus on alerts sent by large data systems, rather than data reports on their functionality. They should spend more time learning the popular and practical details, such as what data the system collects and how to make it easier to export data information (such as CSV files or relational databases), Sabhlok suggested.

"You need to understand the functional data reports and make sure that these reports can answer questions about how to improve your business." "he said.