On Tue, Aug 10, 2010 at 01:26:48PM +0200, Hannes Frederic Sowa wrote: > Do you have insights on how windows is handling the case with RA-flag > other/managed set in the response? Some of the windows boxes in our lab, where there is no RA present*, Solicit periodically. They do not actually reach a long fallback, but rather they appear to fallback for a period and then suspend - on some cycle they return to initial Solicits (setting 'RD', retransmit delay, back to zero and building back up). I have not closely watched the time delay characteristics of this cycle, but they appear to try solicits for about 1.5 minutes every so often. There are no Information-Request messages in these bursts. I don't know what happens where O=0 and SLAAC completes (if the clients emit Information-Request messages without RA-prompting or not). I harbor a suspicion that these 'long cycles' are tied to some system level process, something like Teredo, that probes the lower layer system to ask if native IPv6 access has become available. These probes may cause the system to request configuration. I don't have any justification for this suspicion. * To my embarrassment, we have not yet deployed IPv6 in our testing lab. The lab was built at the last minute after a major catastrophe and I have not yet had time to redesign for IPv6 (we only use IPv6 in production!). It only needed to clone the failed systems, which at that time did not have DHCPv6 deployed (only IPv6 was). -- David W. Hankins BIND 10 needs more DHCP voices. Software Engineer There just aren't enough in our heads. Internet Systems Consortium, Inc. http://bind10.isc.org/
Attachment:
pgpC5Pvb0VDV8.pgp
Description: PGP signature
