|
|
 |
|
|
 |
 |
|
|
|
FAQ: Fail-safe systems and emergencies
-
What if your connection to the Internet fails?
Sanders Hosting is connected to the Internet through several T3s and OC3s to other networks and backbones.
It is very unlikely that all connections from different providers will fail at the
same time, meaning that our servers are always connected to the Internet.
-
What kind of servers do you have?
For more information on our servers, check
the Servers FAQ.
Of course, if you need far more power than what we have, we can get a few extra servers.
Special pricing arrangements for this is done on an individual case basis.
-
What if power goes out?
That's no problem. All servers and networking equipment have battery backup systems and our facilities are connected to two different power companies.
We also have an industrial strength power plant.
-
What if a server's hard drive fails?
With other webhosting companies (over 99% of all), if a hard drive fails, it would mean that:
- Your website can be offline for a day or more while the webhost's owner
goes to buy a new hard disk (this mostly happens on small webhosts, and believe
us, most of them are).
- If a replacement hard disk is close, it'll take at least one hour
to install the operating system and optimize it. Then comes the task of configuring
secondary services (webserver, FTP server, etc), which can take over an hour.
- Most webhosts do backups of user's sites once per day. Depending on how many customers,
it might take several hours to restore all backed up data. But this means that you will
lose all data that was stored on your account since the last backup was made
(imagine losing web sales data of an entire day!).
You might wonder: how often can a hard disk fail? The answer is: often. The hard disk
on your personal computer is not under use very often, so it lasts a long time (while
reading this, your hard disk is probably doing nothing). On an Internet server, this is not the case.
Everytime a visitor requests a page or a file from your website, it has to be read from the hard disk
(unless the file is already on the server's cache memory).
This can happen hundreds of times per second. A solution to reducing hard disk usage
is to increase RAM on the server. The first time a file is requested, it is read from the server's
hard disk , but it then stays on the server's RAM. The next time it is requested, the
file is pulled from RAM and not from the hard drive.
On Sanders Hosting, we have various solutions to prevent hard disk failure:
- First, the server has far more RAM than it's needed, thus taking a lot of
work off the hard drive.
- Two, they are kept extra cool by powerful internal cooling systems. The cooler it runs, the longer it lasts.
- Third, and most important, all data is stored on
RAID storage devices so the probability of downtime
due to hard drive failure is reduced to literally 0%.
This is one very important thing you should consider about a webhost's servers.
And believe us, these long downtimes can happen.
|
|
|
|
10 May 2004
