Load balanced servers
Multiple web, application, or database servers can be configured to work together behind a load balancer. The load balancer decides where to direct incoming traffic, first verifying that the server is available. Utilizing a load balancer as part of your solution ensures that if one of your load-balanced servers is taken offline, your site remains available.
Redundant network devices
Having a true high availability environment means that you don't have a single point of failure in your configuration. Relying on a single device such as one firewall, one load balancer, or one power supply is a possible failure point. Adding a second, redundant device ensures that if one fails, another one is already running.
Clustered database and application servers
Multiple servers can act as a single database or application resource. One or more standby servers are in place so that if the active one fails, another is ready to take its place. There are various options for setting up this configuration on Linux servers and Windows servers, including: database mirroring, DRBD clustering, Red Hat clustering, MS SQL clustering, and Exchange clustering.
How INetU is taking care of it
- Senior-level system administrators are available 24x7x365 so that systems can be tested or transitioned during off-traffic hours.
- With years of consultative experience and a devotion to the business goals of our clients, we will recommend the right solution that is custom-tailored to your needs.
- INetU has a solid history helping clients achieve high-availability success at every level, from configuring load balancers to setting up clustered servers.
Example Scalable Solution
By firmly establishing a web layer behind a load balancer, you increase the number of concurrent visitors your site can handle. Each dedicated web server adds to your site's capacity. Also, as your site grows, you can continue to add dedicated servers behind the load balancer without having to take your site offline. This solution can be easily transformed into the Database HA Solution below when you are ready to firm up that layer with redundancy.
Example Database High Availability (HA) Solution
Setting up a database cluster provides you with redundancy on the database level so that if you sustain a failure on one of your dedicated servers, the other database server can continue functioning and serving data. Database clustering is handled differently in a Windows/MS SQL environment versus a Linux/MySQL environment and our sample depicts a Microsoft SQL Cluster utilizing a shared storage array. The addition of a dedicated backup server makes it possible to easily archive backups for each layer of your infrastructure. This solution allows you to have HA on both the Web and Database layers, and allows for growth into the Enterprise HA Solution below.
Example Complete HA / Enterprise Scalability Solution
Redundant firewalls and load balancers add high availability to your network device layer, meaning that any single device can be removed from this configuration and your site will remain online. With a firmly-established application layer, every component of your infrastructure is scalable, so as the demand on your site increases, you can target upgrades and additional dedicated servers directly to the areas that are causing a bottleneck.
