Load Balancing Brings High Availability
While we all want our servers to be online as much as possible, many infrastructures out there flat-out rely on persistent uptime. Healthcare systems, high-revenue businesses, e-commerce — sites like these depend on consistent and constant accessibility. The smallest hiccup that disrupts user access can cause major issues.
More Servers are Better Than One
It’s an undeniable fact — servers do go down. The best means to work around this disruption is to add replication to your content by hosting it on multiple, redundant servers.
The more servers you add to your infrastructure, the less chance of downtime. Simply, if one server were to go down, then other server(s) would be able to continue serving your data.
The most optimal hardware redundancy results from using multiple servers in different datacenters and across different infrastructure providers. However, very few sites and applications require this level of redundancy.
Alternately, when you run redundant, multiple servers from the same datacenter and from the same provider, you’ll achieve faster speeds and quicker access by which to sync your servers together. You’ll also acquire more tools to help make the high availability process easier.
One such tool is Linode’s own load balancer — called a NodeBalancer.
Load Balancers
The hardware that sits in front of multiple backend servers comprises a typical load balancer and serves incoming requests to content located on those backend servers. When all servers are in the same datacenter, they can communicate via the private network there, which for most providers (Linode included) doesn’t count against your transfer quota and serves data at much quicker speeds.
When a load balancer is in use, visitors will be directed to this upstream server (where the domain name would follow to the IP address of the load balancer, not a back-end server). Then, the load balancer would pass the request to a backend server that is currently running, bypassing any servers that aren’t responding (because of system failure or maintenance).
The live server selected depends on what sort of algorithm the load balancer is using. One algorithm passes requests to the server using the fewest current connections; another would pick a server in a round-robin order. Selection methods can vary, but all do the job.
Linode’s NodeBalancer
Rather than having to set up a load balancing server on your own, you can subscribe to Linode’s NodeBalancer service, where we do all the work for you. Our NodeBalancer functions completely through our Manager interface, and requires only the IP addresses of the backend servers you wish to balance.
You can set extra options, including choosing a balance algorithm, from within the Manager. Full how-to documentation is provided. Our customer support team is also available to help at all times when needed.
With our NodeBalancer, you don’t have to worry about what would happen if a server fails. We’ve gone the extra mile and built automatic failover into every NodeBalancer. If it were ever to go down, another NodeBalancer is immediately ready to manage your content.
On your way to achieving high availability, load balancing becomes essential. And Linode’s NodeBalancer is the catalyst to get you there. You can learn more about our service and how to reach high availability at http://linode.com/nodebalancers.
If you have any questions or comments about Linode’s NodeBalancers, feel free to share below, tweet us, @Linode, or contact our 24/7 support: http://linode.com/support.
