One of the most important hardware components in a hosting server is the hard drive. This is where the physical data is stored.
There are two main types of hard drive:
- Traditional (Spinning) Hard Drives
- Solid State Drives (SSD)
The main difference between them is that the traditional hard drives store data on a spinning data platter whereas the SSD drives are 100% electronic components – there are no moving parts. The SSD Drives are used in our SSD Hosting and SSD Cloud Servers.
In order to maintain high uptimes we as a hosting company need to ensure that there is minimal downtime, especially due to hardware failures such as failed hard drives.
There are several ways we can achieve this and the starting point is to only use high quality, enterprise grade components such as hard drives.
The main failure points in a server are the hard drive and the power supply. Therefore all of our servers have dual power supplies and multiple disks in a RAID Array.
RAID stands for Redundant Array of Inexpensive Disks and there are different variations of RAID configuration. This just means that data is stored across a minimum of two disks using a RAID Controller to reduce the possibility of data loss due to hardware failure – basically the aim is that data is always protected by ensuring that there is no single point of failure, i.e. data is always stored on a minimum of two hard drives.
Taking this further, there are different forms of RAID. For example RAID 1 stores data on two disks exactly the same. This is ideal where data loss is unacceptable, but it performs slowly.
Here at Pickaweb all of our shared hosting servers use RAID 10. This is the optimum solution which balances speed with high availability of the data.
In fact with RAID 10 your data is stored across a minimum of FOUR hard drives, so one hard drive can fail and the server can continue working perfectly with no data loss while the failed disk is replaced.
That’s really important because it means we can keep a high uptime record and not have to worry about a single disk failing.
However, that failed disk still needs to be replaced and that’s where the hot swap feature comes into play. By using hot swap hard drives the faulty hard drive can be removed and replaced in a live server WITHOUT the need to power down the server and replace the failed drive.
Once the failed hard drive is replaced by the new hard drive the RAID array detects the new drive and copies data over to it and once it is fully configured it adds it to the RAID. This results in zero down time.
Now it’s actually a bit of a misnomer to call it a hot swap hard drive because the hard drive is exactly the same as any other hard drive. The only difference is that the server is setup to accept small hard drive caddies which hold the hard drive in place and which can be removed when required.
On top of the Enterprise Hard Drives, the RAID 10 configuration and the hot swap Hard Drives we also back our servers up regularly to ensure that your data is always safe.
Finally, it should be noted that with our VPS servers the disks are configured in a high availability configuration where data is stored on several disks across several servers. This dramatically reduces the possibility of downtime due to hardware failure up to the level of a complete server failing which is an unlikely event.
We hope you have found this technote useful on what is a hot swap hard drive.
Thanks for reading and leave your questions below to keep the conversation going.