Especially during periods of particularly high traffic, you need a way to balance the demands you place upon your servers and network architecture. Too much traffic traveling along any one path can be disastrous – like an overloaded elevator, or a highway during rush hour. As an administrator, you know this; chances are you’re already aware of a few load balancing methods you can use to keep your servers running healthily, even during high-traffic periods.
Today, we’re going to take a look at one method in particular: Global Server Load Balancing, or GSLB for short.
What IS GSLB?
With Global Server Load Balancing, server loads are managed by splitting traffic through the use of DNS and geographical locations. A user connecting in Dubai, for example, will be sent through to a different server than one connecting from New York City. Of course, it can do more than just manage incoming requests, as well.
What Is GSLB Used For?
By and large, the most popular use case for GSLB isn’t even directly user-related: it’s extremely popular as a disaster recovery solution. On some level ,this makes sense – configuring a DR solution with GSLB means it will automatically forward traffic to the least-stressed data center.
Rather than serving to mitigate traffic demands, it instead is used as a tool for detecting when an active site fails, at which point it automatically diverts any incoming requests to a standby site.
This offers a number of advantages over traditional disaster recovery, as it can automatically be used to address platform and server outages, preventing downtime – something that’s especially critical for high-availability websites and applications.
GSLBs also see a great deal of use in regulated industries such as healthcare, finance, and telecommunications. One of the challenges of operating in a regulated vertical is the fact that regulations vary by country and region. By using a load balancer to customize the content a user sees, a regulated organization can better ensure they remain compliant.
Along that vein, a GSLB can be used for marketing purposes as well, delivering localized content to users who visit a site based on their location. Someone visiting a big box retail website from Canada, for example, could be directed to the Canadian site, while someone visiting from Germany could be served content in German. In both cases, people receive information that is more personalized and far more relevant to then. ‘
Finally, delivering content from a server that’s close to a user ensures a better overall experience, allowing for a better-performing website. Not bad, right? Of course, at this point, you’re probably wondering one thing – why haven’t we talked about GSLBs as tools for managing traffic surges?
The reason for that is pretty simple…because they aren’t necessarily all that good at it, at least not when operating in solitude. See, the issue here is that by directing a user in one city to a data center in another, you may be reducing server load. At the same time, you’re increasing the latency for that user, upping their response time and potentially ruining their experience.
As such, GSLB is best-used to enrich your user experience, assist in disaster recovery, and enhance performance. While it can be used for high loads, that doesn’t mean it should.