The old- fashioned manner are still practiced by several businesses and IT sectors. These old approach have some drawback which the present day data have to deal with. Some of these drawbacks are discussed in coming section. However, Bhavana (2015) states that there are few drawbacks which are faced by the data during storage and processing. These drawbacks are also discussed further.
Drawback of traditional approaches
Edelle (2016) claims that there are several drawback which traditional approaches have to deal with. These drawbacks are stated below:
The traditional approach deals comparatively more with the structured data, and storing and processing them requires great cost, apart from this, in traditional approach the data was stored in internal hard-disk which was the other factor for increasing the cost requirements.
Traditionally, the data was present in small amount and managing them was pretty easy. However, the drastic increase in data over last few decade resulted in requiring more space for storage and traditional database fails in providing it.
3. Time consuming
Reading of data from the big dataset requires a lot of time in traditional approach, however this time can be reduced with the help of parallel processing.
Drawbacks of conventional approach
Bhavana (2015) states that there are several other challenges that are faced in storing and analysing the data.
As the data is increasing since last few decades, there is a great demand of storage. These demands resulted in improving the storage capabilities of hard disk. However, the speed of reading data from derives are still same i.e. 100 MB/sec. This leads to take large time while reading of data. In order to have proper time utilization, multiple disk were read in parallel at the same time. This resulted in providing shared access to data storing up to 1 terabyte of data. But, providing shared access may many time cause hardware failure. To avoid this, Hadoop’s file system can provide better solution by providing the feature of automatic failure management (White, 2012).
Since, the data is present at multiple sources which needs to get combined, doing this is a great challenge and this can be easily done the popular feature which Hadoop provides i.e. Map Reduce (White, 2012).