Redundant Array of Independent Disks, or RAID, is a way of keeping content on multiple hard disks concurrently. A RAID could be software or hardware based on the HDDs which are used - physical or logical ones, but what’s common between them is the fact that they all work as a single unit where data is kept. The key advantage of using a RAID is redundancy because the data on all the drives will be the same all of the time, so even if one of the drives fails for whatever reason, the info will still be present on the remaining drives. The general performance is enhanced as well since the reading and writing processes could be split between different drives, so a single one will never be overloaded. There are different types of RAIDs where the capabilities and fault tolerance can vary depending on the exact setup - whether info is written on all the drives real-time or it's written on one drive and then mirrored on another, the number of drives are used for the RAID, and so on.

RAID in Shared Hosting

All of the content which you upload to your new shared hosting account will be saved on quick NVMe drives that work in RAID-Z. This configuration is built to work with the ZFS file system which runs on our cloud Internet hosting platform and it adds an additional level of protection for your site content in addition to the real-time checksum verification that ZFS uses to ensure the integrity of the data. With RAID-Z, the info is saved on a couple of disks and at least 1 is a parity disk - whenever info is written on it, an extra bit is added, so in the event that any drive stops working for some reason, the stability of the data can be verified by recalculating its bits in accordance with what is kept on the production disks and on the parity one. With RAID-Z, the functioning of our system will never be interrupted and it will continue operating effectively until the problematic drive is replaced and the info is synchronized on it.

RAID in Semi-dedicated Servers

The information uploaded to any semi-dedicated server account is kept on NVMe drives that work in RAID-Z. One of the drives in such a setup is used for parity - whenever data is cloned on it, an additional bit is added. In case a disk happens to be problematic, it will be taken out of the RAID without interrupting the functioning of the Internet sites since the data will load from the remaining drives, and when a new drive is included, the information which will be cloned on it will be a combination between the info on the parity disk and data saved on the other hard drives in the RAID. This is done to ensure that the info that is being cloned is correct, so as soon as the new drive is rebuilt, it can be integrated into the RAID as a production one. This is an additional guarantee for the integrity of your data since the ZFS file system which runs on our cloud web hosting platform compares a unique checksum of all the copies of the files on the separate drives to be able to avoid any probability of silent data corruption.