Latency and Why It Matters...

As our personal and business life increases their reliance on data, the ability to move and exchange data is compounded to impressive rates. Our reliance on online gaming, video, VOIP, and other applications makes latency and bandwidth issues a hot topic.

Some common networking terms are listed below:

1. Network speed is generally defined as how much data can flow through your network in one (1) second.
2. Latency measures the span of time it takes a data packet to get sent from a source to a destination, typically measured in milliseconds (ms). Also referred to as the ping rate.
3. Bandwidth is the maximum throughput of a network. Basically, how much data can fit through the pipe?
4. Throughput is how much data can be transferred in a set period.
5. Packet Loss – when a packet of data fails to reach its destination.

In everyday applications latency or lag isn't noticeable, but with latency-sensitive applications, video streaming, or constant feed latency issues can become very frustrating and problematic for the user. For latency-sensitive applications and streaming, data is like a boat in the ocean. Sometimes the ocean is calm, and the boat takes the optimal path with no delays. But sometimes the boat will come across turbulent weather causing the boat to take a less-than-optimal path to reach its destination. Sometimes there are many boats trying to take the same path to reach a similar destination causing congestion in the flow of traffic. The storm is like the routing and processing delays that are not under your provider's control.

There isn't much you can do regarding latency caused by distance as you are constrained by the type of cable and distance. The only option is to utilize a data center closest to your location. When planning your network architecture keep in consideration that every time your data passes through a device, it causes a slight delay as the device decides where to send the packet of data. Additional delays can be caused by firewalls and servers as they receive the request, process, and send a response. Those small delays can add up to a very noticeable and frustrating delay.

Our talented network team is always working to improve the OVHcloud® network and increase the speeds at which your data can travel. To build trust with customers we provide a "Looking Glass" to our network. This site allows you to:

1. Traceroute any destination IPv4 and IPv6 address
2. Show the route for any destination IPv4 and IPv6 address and details
3. Show the route for any destination IPv4 and IPv6 address using a graph

Even with all these improvements, we need your help to pinpoint areas where we can improve.

If you notice latency issues, please send the following data or screenshots to our CST team via the "Need Help" dropdown in your OVHcloud manager.

Reporting the Issues:

1. When did this issue first start?
2. Has there been any recent changes?
3. Source IP
4. Destination IP

At OVHcloud, we suggest using some of the tools listed below to help collect the data needed to allow our network team to dig in and address the issue.

Tools to Collect Data:

1. Iperf3 - https://iperf.fr/
2. >MTR – typically pre-installed on Linux
3. WinMTR
4. JMeter - http://jmeter.apache.org/
5. Wireshark
6. Traceroute

Also note that different operating systems require different troubleshooting methods, if you are using a Linux-based system you can perform the test directly through Terminal. When testing in Windows you will get skewed results unless you boot in Rescue Mode Pro. Not booting in Rescue Mode Pro allows background processes that will use bandwidth and not provide accurate data. You may need to adjust your firewall settings to utilize certain tools and/or test for issues.

If you would like to learn more about the OVHcloud network, please visit us on our Network page.