Hi Pedro, On Fri, Oct 30, 2026 at 03:03:58PM +0000, Pedro Ferreira via NANOG wrote:
We are researchers at UFMG and FURG (Brazil) studying how networks use the "never via route servers" option in PeeringDB and how IXP route servers apply it.
If you operate a network, we would appreciate 3 minutes of your time, whether or not your network has the option enabled: https://forms.gle/etf7ezkd5vWgPKoA6
Participation is voluntary and anonymous. We collect no personal data, identifying your network is optional, and results will only be reported in aggregate.
For questions or comments, contact us at: Pedro Ferreira pedroferreira@dcc.ufmg.br mailto:pedroferreira@dcc.ufmg.br , Italo Cunha cunha@dcc.ufmg.br mailto:cunha@dcc.ufmg.br , Pedro Marcos pbmarcos@furg.br mailto:pbmarcos@furg.br
In the survey you ask: """Which expected effects of the "never via route servers" flag motivated your network to enable it on PeeringDB? (Select all that apply)""" But to me it seems the set of multiple-choice answers does not contain the answer which motivated the creation of the 'never-via-routeservers' flag. The primary motivation to create this flag (and use this flag) is that some networks wish to avoid their routes being part of route leaks propagated via route servers. The survey suggests some related answers: * Route leaks through my network do not spread via route servers * Fewer route leaks reach my network * My routes are dropped by route servers But I think the real answer is a mixture of the three: "route leaks containing my network do not spread via route servers" Kind regards, Job