Today's networks (data centers, ISPs, and enterprises) all run many hardware middleboxes and software network functions. The performance of these components is critical for network management and user experiences. However, when performance problems happen, it often takes a lot of human efforts to instrument and debug these systems especially when many network functions interact with each other and with traffic dynamics.

We are a group of researchers from Harvard and AT&T who are building new diagnosis tools that can automatically find potential root causes for performance problems in network functions (especially those tail latency problems that are hard to diagnose).

We're reaching out to the NANOG and other Network Operators' Group communities to understand the performance problems you have seen and your needs on diagnosis tools. This will greatly help us to incorporate your needs in our tool. We plan to release our debugging tool for the community to use.

We greatly appreciate your help in filling out the survey below. It should take less than 10 minutes to complete. The survey and the collected data are anonymous (so please do *not* include information that may identify you or your organization). All questions are optional, so if you don't like a question or don't know the answer, please skip it. In the survey, we use network functions as a general term to represent both hardware and software middleboxes. A summary of the aggregate results will be published as a part of a scientific article later this year.

We would also be extremely grateful if you could forward this email to any operator you know beyond this mailing list.