Dear All, In the framework of research projects, we are working on techniques to reduce the impact of planned maintenance operations on the routing protocols and the packet forwarding (see e.g. discussions on ordered IGP convergence in RTGWG at IETF). Several studies of ISIS or OSPF data have shown that maintenance operations are very frequent in large ISP networks. We would like quantify the number of maintenance operations that occur in large networks and would appreciate if you could answer the following survey (please reply directly to me, I'll summarize to the list) : Network size : - approx # routers - approx # links - approx # eBGP sessions Maintenance operations : - How often do you need to completely reboot each of your routers ? [ once per day, once per week, once per month, never,...] - How often do you change the IGP metric of each of your links ? [ once per hour, once per day, once per week, once per month, never, ...] - How often do you manually shutdown each of your intradomain links ? [ once per hour, once per day, once per week, never, ...] - How often do you need to stop and restart each of your eBGP sessions ? [ once per hour, once per day, once per week, never ...] - Do you use special operational procedures to perform those maintenance operations to reduce their impact on routing and packet forwarding ? [ please provide comments if any ] Thanks for your help. Olivier Bonaventure -- CSE Dept. UCL, Belgium - http://www.info.ucl.ac.be/people/OBO/