High on the list of network engineering nightmares is a business critical process failing because it didn’t receive adequate bandwidth. With applications, users and data competing for bandwidth, how do you assure your business-critical applications are getting the bandwidth they need for optimum performance? The best approach is a combination of network QoS (Quality of Service) policies and bandwidth utilization monitoring.
Related Article: How Application Performance Monitoring Keeps the Cloud on Track
QoS and Bandwidth
Defining QoS policies in your network infrastructure allows you prioritize or guarantee the allocation of network bandwidth to different applications or users. So QoS is key to assuring that mission critical applications or business processes get the bandwidth they need. QoS can help you avoid the nightmare of having a business update process fail because a group of users are streaming a sports event. QoS policies can take various forms such as:
- Policing: Ensuring a business critical application gets a pre-defined minimum level of bandwidth or that non-critical applications are limited to a maximum.
- Shaping: Queueing non-critical excess packets for transmission over time to prevent burst traffic spikes that can impact other applications.
- Priority queuing: allocate bandwidth to prioritized applications (like VoIP), before doling out to others.
But, depending on your environment, it can be difficult to fully define a flawless bandwidth utilization environment. That streaming event may be happening in the board room or IT might not be in the loop regarding a new process/application that sales has rolled out. So you need a way to proactively monitor bandwidth utilization so you can be alerted at the first sign of developing bottlenecks, violations or new bandwidth hogs. Best practice would be to implement a network traffic flow monitor.
Network Traffic Flow Monitor
Flow monitors analyze bandwidth usage to provide useful insight to applications and users that are top traffic consumers. They can often be configured to alert you to bandwidth hogs or bandwidth deprived critical applications. A “flow” is a contiguous series of packets that share the same characteristics like source and destination IP addresses, source and destination ports, IP protocol, Type of Service etc. If your network hardware supports the capability, flows are collected and forwarded to a flow monitor. There are vendor specific implementations of flow collection, so your flow monitor should be conversant in whatever format your network equipment vendor(s) support.
When selecting a flow monitor solution here are some key characteristics to look for:
- Support multiple, vendor specific flow collection standards (NetFlow, SFlow, J-Flow, IPFIX, NetFlow Lite, etc. ).
- Pre-packaged and customizable reports and dashboards that let you identify top senders, receivers, conversations, ports, and protocols consuming network bandwidth.
- Real-time alerting to problematic trends, bottlenecks and policy violations.
- QoS verification through pre and post policy implementation views.
- Automatic discovery and configuration of flow enabled devices to save time and ease configuration
With the combination of QoS and flow monitoring you can be pretty confident that you won’t be called in to the executive suites to explain a major bandwidth utilization issue affecting the business. And, of course, Ipswitch has a robust flow monitor plug in to WhatsUp Gold that you should check out.