3 Network Problems That Affect the Cloud App End-User Experience
by October 23, 2017

Filed under: Cloud Computing, Performance Monitoring

When you’re using the cloud, the network is a crucial part of every application. There isn’t anything local; all access is over either the public internet or a direct connection to the cloud. While the cloud provider can ensure adequate resources for performance on their backend, they have no control over the rest of the application delivery path, let alone how poor network performance impacts end-user experience. To understand that, you need to use network monitoring software and keep an eye out for these three big network problems that can negatively affect cloud application users.

1. High Network Latency

If it takes a long time for data packets to get from the user’s device to the application server in the cloud, the user will perceive delays and slowness in application performance. Sometimes latency is due simply to the distance the data is traveling, and you might be able to improve performance by asking the cloud app vendor to switch you to a server or data center in a different region closer to the users.

Latency can also be due to networks that are very busy. This often happens when you’re using the public internet to connect to the application, which is where MPLS and SD-WAN network considerations come in. If the latency is due to delays on your internal network, you need to check why the network is so busy—you may need to block users from non-work-related sites.

Another cause of latency is low bandwidth, the limit on how much data the network can handle at one time. If your internal network’s bandwidth isn’t adequate, and you’ve blocked all non-work usage, you may need to speak to your network provider to increase the bandwidth. Also remember the importance of monitoring capacity, which can provide more in-depth detail about usage than bandwidth. In addition to higher speed bandwidth, consider options for higher quality bandwidth with less packet loss and retransmission. Business-class internet services and dedicated circuit networking can alleviate these issues.

2. Network Misconfigurations

The way applications and the network behave depend on how they’re configured; errors in these settings can lead to noticeable performance degradation. In particular, DNS lookups are used to identify which servers to connect to, commonly based on geography. If the wrong DNS server is used, users may not be connected to the closest server and experience poor performance as a result.

3. Network Intermediaries

In corporate settings, users rarely connect directly to outside applications. Their requests pass through proxies, firewalls and other devices that mediate access and control network traffic. Misconfigurations in any of this software can lead to delays and perceived application slowness.

While intermediaries can add delays if they aren’t properly configured, sometimes you can use intermediaries to improve performance. Intelligent caching can reduce the time needed to access sites by storing frequently accessed files that rarely change.

Some major network problems for cloud end users

Use Monitoring Tools to Identify Network Issues

Your cloud provider can’t help you identify these issues, so you need to use third-party network monitoring software. The monitoring needs to be able to see what happens to your traffic both on the public internet and once it reaches the provider so you get an end-to-end understanding of where any performance problems are coming from.

To get end-to-end visibility, the network monitoring software you choose should be able to monitor the system behavior starting at the user’s device through to the cloud application and back. It should work whether the employee is at company headquarters, a remote office or working from home, whether their connection is via WiFi or a WAN/LAN, and whether the user connects via a private network.

Synthetic transaction monitoring gives the best measure of the end-user experience. Low-level measures of network performance don’t tell you how the overall application is performing for users. By simulating exactly what the user does, the result isn’t an estimate of performance; it’s the actual performance. Synthetic transactions aren’t stopped by firewalls on the cloud provider’s side; you’ll cover the full transaction.

Use Monitoring Tools Designed for the Cloud

You can try to leverage your existing network performance monitoring tools, but they often don’t work well when you need to monitor connections to the cloud. Tools that weren’t built for the cloud usually can’t get a detailed look at what’s happening outside your firewall. But when you’re using applications in the cloud, that’s where most of the action takes place.

Tools that are designed to monitor network connections to cloud applications and that use synthetic transactions give much more detailed and specific insight into network issues that impact the end-user experience, wherever the problem originates. Investing in new technology designed to work with new cloud-based systems can reduce the effort needed to analyze network problems. Modern monitoring tools like AppNeta’s generate data that backs up requests for support from the vendor to get issues resolved more quickly.

With the right modern network monitoring software and tools, you’re able to focus on what really matters: making technology work for people, rather than aiming for performance metrics as an end in themselves. Request a trial or demo from AppNeta today and see how our network monitoring software can help you.