High Availability and Business Continuity
By controlling traffic during load spikes, crashes or server maintenance, HAProxy application load balancing solutions ensure that online services and applications are always available.
Absorbing load peaks
Occasional or recurring load peaks slow down response times, reduce quality of service and even make Web access applications or services unavailable. To compensate for increases in traffic, HAProxy load balancers balance traffic intelligently based on the servers’ availability and/or relative capacity.
Acting at the application layer, the load balancer does not “blindly” route to the server that seems to have the lightest load (number of open sessions): it analyzes the content of each query, using predefined algorithms, to determine the best routing.
Rerouting traffic when a server crashes
HAProxy application delivery controllers continuously perform server health checks: in a few milliseconds, they can isolate a malfunctioning server, by interpreting a series of incorrect status codes. And implement temporary solutions completely seamlessly for the user: dynamic switching to reroute traffic, backup servers and http redirection and disaster recovery plan (information about the server crash).
Rerouting traffic when maintenance is being performed
This function to reroute traffic automatically to other servers also makes it easier to perform routine maintenance: simply delete the server or server pool undergoing maintenance and the load balancer automatically redirects traffic to other servers, which can be added on the fly.
This gives IT teams more flexibility; they are not forced to perform maintenance during low load hours, often at night or on weekends, to avoid adversely affecting the company’s activity.