Blog - nysernet

Innovative Campus Connectivity: Insights from Our Network Engineers

Written by nysernet | Jul 10, 2025 4:10:14 PM
What makes a connection reliable?  

That’s hard to say, as no network is 100% reliable. You cannot guarantee that a connection will always work. The key is that the network's users need to be insulated from the problems that occur. The number one thing you need for a reliable network is redundancy.   

Outages are always going to happen (with upgrades, maintenance, etc.), but having a redundant network, such as backup connections or diverse routing, can help you get as close to that 99.99% uptime as possible, even during routine maintenance or unexpected issues.   

With redundancy, if you have one, then you have none, but if you have two, then you have one. Think of it this way, insurance (like health or home insurance) doesn’t erase the problem (hospital visit, house fire, cyber breach) it just reduces the pain. Redundancy is different than insurance, because you have the protection and reward without an incident happening.  

  

Wired versus wireless  

Which connection is better, wired or wireless?  

One is not better than another; it's on a case-by-case basis. Wireless is about convenience and wired is about performance. Certain cases call for wired, while wireless works well for others.  

From a research aspect, (working with, moving, and downloading data) wired connections create a high-performance, more reliable network. However, when manipulating and viewing data, it can work wirelessly. Most campuses use wireless connections for their students, as well as their faculty and staff.  

Fun fact: Several of our members use NYSERNet’s wired infrastructure to conduct research on wireless connections. That says a lot about the critical role wired still plays in today’s high-performance environments.  

  

How does NYSERNet monitor their networks? 

You can pull most of your device performance directly from the devices, but for comprehensive monitoring and performance analysis, consider using these tools:  

  • SNMP (Simple Network Management Protocol):  Collects real-time data from networked devices (like switches or firewalls). Great for tracking interface utilization, device uptime, and hardware status.  
  • NetFlow:  Monitors network flow data to see what types of traffic are moving across your network. Helpful for identifying unusual spikes in usage or application-specific bottlenecks.  
  • iPerf:  Generates traffic between two endpoints. Perfect for benchmarking throughput or stress-testing segments of your network.  

Use SNMP for continuous monitoring, NetFlow for usage analytics, and iPerf when you need to push your network through a controlled test. By combining the three, you can create a proactive approach to network performance management.  

  

Understanding Latency and Jitter  

Latency impacts performance in two key ways: 

  1. The total time it takes   
  2. How much that time changes (jitter).   

Think of it like this, the speed of the car represents how fast the data is moving. Jitter is how often the car has to slow down or speed up, which makes the trip less predictable.   

Latency is about travel time. The lower the latency, the more direct a connection. In data centers, where there is little distance between the connections, low latency is very easy to achieve.   

When you're transferring files, latency matters the most. Imagine a large pipe filling a water tank. The length of the pipe is your latency. Once water (data) starts flowing, the size of the pipe (bandwidth) determines how much it can move. But if it takes too long for the network to get acknowledgments back, it can slow down the entire process, making it harder to fully use the bandwidth.  

Again, it’s easier to do it in a data center because there is less distance for the data to travel. Low latency can fill the pipe better.   

Jitter becomes important when you depend on predictable latency. Lots of jitters can cause echoes on video calls, live streaming, esports, and many other real-time applications. Low jitter is key to those real-time applications because it creates a consistent flow.  

  

Can connectivity inherently provide cybersecurity protection? 

Connectivity does not directly provide more security to your networks. However, the R&E network benefits from not seeing the same attacks a commercial network does. At NYSERNet, for example, we offer dedicated connectivity paths that bypass the commercial internet entirely, reducing exposure to common attacks. 

Depending on your use case, specialized connectivity solutions can indeed enhance your overall security posture. It’s not the connection itself; it’s how and where it connects.  

  

How to decide between performance and cost? 

Weigh the cost of having good connectivity versus not. In the case of a large outage, you may be facing an even larger cost than just paying more for connectivity. Think of selecting connectivity like choosing an insurance policy. Sure, you can save money by going with the lowest-cost option, but what happens when something goes wrong?  

In practice, this often means choosing between:  

  • Good performance at low cost, or  
  • Better performance with a slightly higher cost, enabling failover, scalability, and consistent throughput.  

When faced with an incident, the cost of not having something is often far higher than the cost of having it.  

  

It doesn’t have to be perfect. 

Connectivity may never be 100% reliable, but with the right strategy, redundancy, smart monitoring, and tailored design you can come close. It’s not about eliminating risk; it’s about planning for it, minimizing its impact, and keeping your network ready for whatever comes next. 

At NYSERNet, we believe reliable connectivity is about building resilient, high-performing infrastructure that supports the education and research institutions we serve. Our network engineers work behind the scenes so your campus can stay focused on what matters most.