One of the most common type of system failures in web hosting today is a hard drive failure and since we use RAID-10, that results in a degraded RAID-array and that is borderline worst case scenario.
This means that the drive must be replaced or data loss is highly likely. It’s our first line of defense. We also use RAID-10 for performance reasons.
The simple fact is that hard drives fail and is not a matter of if but when! Here is an article by PC World on a study that was done on hard drives by Carnegie Mellon University & Google.
The good news is…
Thanks to the experience of our management team and the excellent team that we have at GreenGeeks, the pooled knowledge has allowed us to combine work experience to ensure that customers data is safe and downtime is reduced to an absolute minimum during a failure. We are all about proper procedures & policy. We have replacement hard drives (including other hardware like RAM, CPU’s, Server Chassis, etc) on standby in the event of a failure, we’re able to quickly replace the affected hardware.
In the event that there is a file system / hard drive failure, there are likely two scenarios that would occur:
- Bad Drive (best case scenario) – We simply replace the drive and the hardware RAID card will re-build the new drive with the lost information from the data on the other drives and in a few hours the RAID-array will be repaired all while your sites are still on line, with no data loss. Total downtime may be less than 1 hour.
- File System Corruption (worst case scenario) – If a RAID card fails, this can result in a failed a corrupted file system and a full system restore from backups will need to take place. Restoration time can take a few hours, however we have servers standing by and we never use the same hardware components again.
When there is file system corruption, we attempt to fix it within a reasonable restoration time, if this is not met we proceed to restoring the service from backups.
About our backups!
We back up all of our servers, every night and retain the data for 24 hours in the event of a total failure requiring restoration from backups. Our disaster recovery effort! We are also in the process of making our backups redundant as well.
Bottom line is that it’s not a matter of if, but when and being proactive/prepared for the worst is what is going to ensure that data is not lost like it is with other hosting companies (won’t name names!). Rest assured, we take our business very seriously and that’s what we’re all about — quality service and no short cuts.
GreenGeeks / Systems Administrator