AnyHosting

I’ve been doing some research to prepare for writing a new article on hosting reliability in general. In particular I’ve been brushing up on geographical failover, as well as routing users to the server that’s going to be most responsive for them (these two problems have a lot of solutions in common).

I am really surprised that there seems to be a lot of general information out there on Global Server Load Balancing (aka GSLB), and not a whole lot about multihomed setup using BGP (aka Border Gateway Protocol; this really just means announcing the same IP from multiple places).

GSBL has recieved a lot of criticism, and I think rightly so (although more balanced views are available). The fundamental idea is to serve different A records to different clients, which isn’t really new, but GSLB also tries to intelligently route clients to the best IP for them (sometimes by sending out probes, which administrators tend to dislike).

The main problem with GSLB is that it relies on the DNS Time To Live setting being honored, but clients and servers just don’t honor this in the real world (much to the chagrin of anyone who’s ever been forced to change IPs on short notice).

Multi-homed BGP announcements, in contrast, announces the same exact IP address from multiple routers in the world. BGP is all about routing, and the shortest route will tend to be the fastest. It’s also fairly self-recovering, although it usually takes a few minutes (generally no more than 5).

GSLB is much faster (pretty much instant) to failover, if the client doesn’t ignore the TTL. BGP works the same everywhere, and is much simpler to implement (as long as you have control over your IPs!).

The gotcha is that your service provider needs to be willing and able to provide this service to you. I know that many do, but it’s not always advertised so it’s hard to get a good gauge of how common this is for smaller sites.

This entry was posted on Thursday, May 17th, 2007 at 11:58 pm and is filed under web hosting. You can follow any responses to this entry through the RSS 2.0 feed. You can leave a response, or trackback from your own site.