It gets easier and easier to create and keep data, but it is always easy to lose data. In fact: some historians believe that this current digital age will make it more difficult to conduct research and find documents because destruction or data loss is so widespread.
When you use a traditional hosting company (not cloud): there is a 100% chance that the disk will crash one day or another! It is just a matter of time. As time goes by, your disk is getting closer and closer to their MTBF (Mean Time Before Failure) measured in hours. SSDs drives (including NVMe, M.2) have a higher lifetime than mechanical HDD drives but they use another metric: Terabytes Written (TBW), which measures the maximum amount of data that can be written before the drive fails completely.
So regardless of the drive type:
Your disks will fail!
Using RAID does not change anything, as this technician in a datacenter discovered:
Despite using RAID 10: two disks in the array failed within a very short period time before they could get replaced, rendering the entire four disks useless! Very rare!Said a happy technician to have something exciting to share
The customers were very happy to be the casualty of this rare event, having lost data, and resulting in a 36 hours downtime:
I have lost one week of emails and work, because the newest backup available was 1 week old. The restore of the server was slow because the hosting company did not put SSDs but re-installed normal and cheap HDDs.Said a clueless customer
There is a reason why Cloud Computing is superior:
- drives do not fail because the industry’s metric is: data durability
- you can switch between SSD and HDD without downtime
- backups can be put in place very easily
- backups are cheap (sometimes less than $/€/£15 per YEAR)
If you run any website: contact us for a cheap backup solution?