At Exinda, we hear a lot about other solutions that also provide packet shaping. This often comes up when one of our customers is replacing a component of their network, such as a firewall, and the product they’re evaluating claims to also provide packet shaping. On the surface this sounds attractive – “I can buy one product and get multiple benefits.” In the right circumstances that can work, but in many cases it results in a lack of flexibility and an inability to provide the organization with the right solution; resulting in support issues and poor service. As a company with over 10 years of experience solving bandwidth management challenges, we’d like to think we’ve learned what helps network admins successfully manage bandwidth.
YOU MIGHT ALSO LIKE: Considering a Bandwidth Upgrade? Think Twice.
So, what are the specific capabilities that are most important to our users?
1. Logical Separation of Bandwidth
The ability to logically separate traffic into multiple buckets and divide the available physical bandwidth between these buckets is important for a variety of reasons. The most common need for this capability is to separate multiple physical locations behind a single ingress point.
We have also seen cases of organizations wanting to separate traffic for different user groups, subnets, etc. One of the best examples of this, are school districts. They have a single ingress point for the district, but need to manage bandwidth to each individual school separately because of physical networking differences, school size differences, etc.
Without this flexibility, the network admin is forced to try to service all logical networking boundaries from one set of bandwidth rules, which may not make sense for the organization.
2. Intuitive Policy Management
The ability to understand how to create policies that result in a particular behavior, probably sounds trivial. However, we often see companies struggling with this. Many packet shaping solutions are not very intuitive. They separate the policy from the bandwidth rules that will be applied. Users are rarely confident that what they are trying to do is what will actually happen.
You could trivialize this as a learning curve, but imagine investing weeks into one Network Admin learning to use a tool and then they leave, or they move into a new role, and the learning starts all over again. You could also conclude that this just an inconvenience, but this separation of the desired of policy and action can lead to unintended results and leave a network exposed.
3. Unlimited Bandwidth Configurations
Network Admins need the flexibility to create the number and combination of bandwidth rules that meet the needs of the organization. Let’s illustrate this by example. Imagine a school district with 20 schools. Between these 20 schools there are 4 unique bandwidth requirements (i.e. schools 1-5 require 1 Mbps guaranteed with 2 Mbps maximum, schools 6-10 require 2 Mbps guaranteed with 3 Mbps maximum, schools 11-15 require 3 Mbps guaranteed with 5 Mbps maximum and schools 16-20 require 5 Mbps guaranteed and 8 Mbps maximum).
Each school is enforcing 10 bandwidth policies (guaranteeing a certain amount of bandwidth to critical learning applications, online testing and VOIP while at the same time choking bandwidth to P2P applications, video streaming and selected social media applications).
Amongst these 10 policies there are 8 unique bandwidth rules. If we overlay the 8 unique bandwidth rules per school with the 4 unique bandwidth rules to the groups of schools we end up with 32 total unique bandwidth requirements.
This is a conservative example; we often see cases where of 50+ (to several hundred with service providers) unique bandwidth combinations are required.
4. Flexibility and Convenience of Percentages
The ability to manage bandwidth rules via percentage (instead of just raw bandwidth) has some significant administrative benefits. It is convenient and it helps future bandwidth rules. Take our previous example where we had 4 different school groups with different overall bandwidth requirements. Inside each of those groups we have policies that also have bandwidth requirements. By leveraging percentages on the individual policies (for example, online testing with receive a guaranteed bandwidth of 10%, with an ability to burst up to 40% if bandwidth is available) we can create one set of policies that can be applied to all schools regardless of the group the school belongs to.
As I upgrade bandwidth, or move schools between groups I don’t have to make any changes to the individual policies. This can save a lot of time and avoid situations where bandwidth is incorrectly allocated because a particular policy didn’t get updated when an upstream change to bandwidth was made.
5. Dynamically Adjust
In the world of Bring Your Own Device (BYOD) it is critical that bandwidth rules adapt as new hosts/devices join the network. It might be important to cap bandwidth for each individual device, or cap the number of devices, or it might be important to share available bandwidth fairly.
Universities provide a great example of where the ability to dynamically adjust is critical. Without dynamic bandwidth adjustments, the network can be held hostage by the first user on the network that starts downloading torrents.