Gerald, I think there may be an issue of terms here. I believe that there is relatively little truly incorrect data in the RADB. By "incorrect", I mean data that would result in incorrect or missing routing. There is a fairly large amount of "obsolete" data that is virtually impossible to track down because it is abandoned. Routes that were transferred from the PRDB are especially likely to be obsolete, but I have stumbled on a number of routes registered to various organization and placed in the RADB when the organization either no longer exists or no longer is connected using that route. This is a less serious problem in that it never results in failed routing. It does have some slight security implications, but I don't see a really significant way to exploit these. The bottom line is that we (ESnet) has been using the IRR to generate portions of our router configuration for several years with very good results. And I still believe that record expiration will be needed some day because the database will accumulate abandoned data and continue to grow until things start to get un-manageable. I have no idea how close this is to happening, but I suspect that it is NOT near. In any case, I think we are in basic agreement. -- R. Kevin Oberman, Network Engineer Energy Sciences Network (ESnet) Ernest O. Lawrence Berkeley National Laboratory (Berkeley Lab) E-mail: oberman@es.net Phone: +1 510 486-8634