Fail-Over Website Hosting

In April 2011, Amazon had a near four day outage of their EC2 cloud. FOUR DAYS!!!! And by one of the major enterprise hosting providers in the world. You HAVE TO spread your risk.


Let's face it. Website downtime is stressful, and costly. Most businesses cannot afford to have their website down, but outages are common place. Hardware can and will fail. Networks can fail. Facilities can fail.

Most website hosting companies offer a 99.9% up-time guarantee. As consumers, we buy into this, and hope for the best. Only when things go wrong, do we start to pay attention, but by then, it's too late.

A 99.9% up-time guarantee still allows for 45 minutes of unplanned downtime in a month (0.1% of 30 days).

When hardware fails, outages are usually longer than 45 minutes, and when their 99.9% uptime guarantee is breached, most hosts will credit you a pro-rate portion of your hosting fees.

Is a $5 credit going to make amends for a 2 hour website outage?


Globi has created a website fail-over hosting environment to ensure your website stays up, no matter what. We do not depend on a single provider, and instead, spread our fail-over network over multiple providers, in different data centres, on multiple networks, in different cities.


Website Fail-over

If there is a problem with the primary server that your website is on, your website traffic will automatically be routed to the secondary server in the alternate data centre.

When the problem with the primary server has been resolved, traffic will automatically be routed back to the primary server.

Remember, your risk is spread over multiple providers and cities. Any single-provider fail-over solution still has a single-point-of-failure. This is not the case with Globi's fail-over website hosting solution.

Email Fail-over Queue

Usually, when your mail server is unable to accept mail, your incoming emails are simply bounced back to the sender.

The Google / Postini fail-over queue will retain all emails that cannot be delivered, until the primary mail server is back online. You'll never miss an email again.

Static Cache Backups

For most of us, having to maintain two separate websites is too much effort. Our caching engine regularly and automatically scrapes your website content as the fail-over solution.

There's nothing for you to configure or maintain. A static version of your website will automatically be available in case there's a problem with a primary server.

For more details, read about the Globi Caching Engine and how it works.

Fully load-balanced interactive Fail-over Hosting

Should you require your website to be available dynamically, just let us know. You will have full access to your second website instance, and therefore can maintain both websites and keep them in sync.

Both sites will be served equally depending on load (load-balancing), and in the event that one site becomes unavailable, only the other website will be served until both are available again.

What's the catch?

There's no catch. There are only a few technical points to observe and consider:

DNS propagation can take 5 minutes. This means that your website may be unreachable for some visitors for up to 5 minutes. For many visitors it will be always reachable. But even 5 minutes of spotty connectivity is way better than 45 minutes or more of complete website outage.

Please note that many common web applications are not designed to work on multiple servers. If you wish to have a fully operational fail-over website, both your websites will need to synchronize with each other. That is why we offer the static cache - this option is more than sufficient for most websites.