Storage Of Large Data

How to store big data. Basic approaches

What's the big data? The answer to this question depends on who and when asked. We'll get a common user: fifteen years ago, an average home computer was computed by several gigabytes, now hundreds and even thousands of hygabytes. More serious example: Boeing Jet sensors generate about 10 TB data from each engine for only 30 minutes. I mean, a plane coming from Moscow, say, to Novosibirsk in four hours will give us about 160 TB data. And it's only from one flight. For dessert, it is possible to calculate how many of the data left to humanity the past Olympics in Sochi: hundreds of athletes and their data, thousands of hours of competition videos, surveillance footage, etc.

Big data are both big problems and great opportunities. We will consider a few typical issues related to Big data.

We will. As we've just noticed, there's a lot of data, and they're constantly growing. This requires innovative devices and algorithms in principle. storage information♪
Speed. The data themselves are almost useless if they are not processed and processed quickly. By the way, speed is a very relative concept, and the fact that for some data is very fast, for others it will be slow.
Heterogeneity. Data can be very different: by importance, rate of renewal, addition, etc. All this requires different storage formats.
Security. Data should not be lost; unauthorized access is also undesirable.

The list could continue, but any problem was reversed. Amazon, known as its online store, only earned about $4 billion on its cloud services in 2013. In 2014, it is estimated that this amount could range from 6 to 10 billion.
There are three ways to store digital data:

Traditional: " Where is it " on disks, tapes, local warehouses, etc.;
In public lounges: from giants like Amazon, Microsoft and Google or smaller companies;
In private " rooms " : a more generic version of the corporate segment; the repository is part of the company ' s infrastructure and is accessible only to its employees.

We will address some of the advantages and disadvantages of these approaches.

ение Maintenance of " self "

Most common to most of us. The information is recorded on local repositories - discs, RAID masses, tapes, etc.

Plus

It's common. Data is always around, and we're safer.
Access speed. In general, the local host can be easily and quickly connected.
Price. Although it might be a minus.

Less

Unreliable. Disks and servers are degraded by physical wear. No matter how reliable a server, he's not gonna protect the data from natural disasters or banal theft.
Access to data. There is no distance, uncomfortable, or at least not always safe.
Large. Its capacity is generally limited. We need to buy new hosts and place them somewhere. What if you want 10 Tb today, tomorrow is only 5 and the day after tomorrow is 50?