Data Compression
Learn precisely what Data Compression is and discover how it may affect your sites as well as the experience of your visitors.
Data compression is the compacting of info by lowering the number of bits that are stored or transmitted. Thus, the compressed info requires less disk space than the initial one, so much more content can be stored on identical amount of space. You will find different compression algorithms that work in different ways and with a number of them only the redundant bits are erased, which means that once the data is uncompressed, there is no loss of quality. Others erase unneeded bits, but uncompressing the data at a later time will lead to reduced quality compared to the original. Compressing and uncompressing content takes a huge amount of system resources, especially CPU processing time, therefore every hosting platform that uses compression in real time must have sufficient power to support that feature. An example how info can be compressed is to replace a binary code such as 111111 with 6x1 i.e. "remembering" how many consecutive 1s or 0s there should be instead of storing the actual code.
Data Compression in Shared Hosting
The ZFS file system which is run on our cloud web hosting platform uses a compression algorithm called LZ4. The aforementioned is a lot faster and better than any other algorithm you can find, particularly for compressing and uncompressing non-binary data i.e. internet content. LZ4 even uncompresses data quicker than it is read from a hard disk, which improves the overall performance of Internet sites hosted on ZFS-based platforms. Due to the fact that the algorithm compresses data very well and it does that very fast, we're able to generate several backup copies of all the content kept in the shared hosting accounts on our servers on a daily basis. Both your content and its backups will need less space and since both ZFS and LZ4 work very quickly, the backup generation will not change the performance of the web servers where your content will be kept.