Efficient and reliable exchange of data is necessary for world-class scientific collaborations. The networks that connect scientific teams need to handle increasing amounts of data and convey it across multiple networks around the world. However, end users only care about end-to-end performance, not everything that happens at points between.
When a network is underperforming or errors occur, it can be complicated to identify and solve the source of the trouble, as problems can occur across multiple networks. Troubleshooting the source of compromised network performance can be time consuming. Local testing will often not flag the cause, as errors can occur anywhere within a network, or anywhere on the path of data as it travels and is transferred between multiple networks. While hard failures are relatively easy to detect and fix, soft failures where a network continues to function but has compromised performance, can go undetected for years.
This is where perfSONAR comes in. perfSONAR automatically tests and monitors networks, looking for sluggish performance. It provides network users and operators with an efficient way to troubleshoot network performance problems, automatically monitor paths of interest, and collect and archive information on network performance. perfSONAR identifies areas of low performance, both by their location on the network and by the time window in which they occur, and flags these trouble spots.
How to effectively use perfSONAR
If you don’t know what your network is capable of doing, you may not know when it is underperforming. ESnet is working with users to better characterize their networks and measure performance to more effectively root out incipient problems that can impair performance. For a guide on how to use perfSONAR to troubleshoot network performance issues, check out our fasterdata knowledge base.