> From: Margaret Wasserman <[EMAIL PROTECTED]>
> > > That's certainly an option. The "new prefix causes several thousand > > > nodes to all attempt DAD at the same time" argument is the one which > > > makes me hesitant to simply support this without further investigation. > > If thousands of nodes are involved, the DAD optimization may not be > enough to fix this problem. Umm? DAD optimization definitely removes the quoted problem. When optimized, DAD is only done on the link local address when a node is attached (or booted) to the network. > Instead, we might need some sort of random delay on the creation of > addresses when a new prefix is received. The random delay is already specified for DAD in the boot/attach process. The delay is less suitable for DAD's done on RA. In optimized environment, addresses are useable immediately after RA. With non-optimized and delayed DAD, we would have undetermined gap after RA, when an address retrieved from DNS might not work. I would like my network to be as deterministic as possible, avoid randomness when not absolutely necessary. -------------------------------------------------------------------- IETF IPng Working Group Mailing List IPng Home Page: http://playground.sun.com/ipng FTP archive: ftp://playground.sun.com/pub/ipng Direct all administrative requests to [EMAIL PROTECTED] --------------------------------------------------------------------
