RE: [dhcwg] dhcpv6-24: randomization delay before invoking DHC

This issue on the flip-flop of the ManagedFlag is already addressed in RFC 2462:

5.5.3.  Router Advertisement Processing

   On receipt of a valid Router Advertisement (as defined in
   [DISCOVERY]), a host copies the value of the advertisement's M bit
   into ManagedFlag. If the value of ManagedFlag changes from FALSE to
   TRUE, and the host is not already running the stateful address
   autoconfiguration protocol, the host should invoke the stateful
   address autoconfiguration protocol, requesting both address
   information and other information.  If the value of the ManagedFlag
   changes from TRUE to FALSE, the host should continue running the
   stateful address autoconfiguration, i.e., the change in the value of
   the ManagedFlag has no effect.  If the value of the flag stays
   unchanged, no special action takes place. In particular, a host MUST
   NOT reinvoke stateful address configuration if it is already
   participating in the stateful protocol as a result of an earlier
   advertisement.

So, it shouldn't need to be restated in the DHCP specification?

- Bernie

-----Original Message-----
From: Thomas Narten [mailto:narten@us.ibm.com]
Sent: Wednesday, May 08, 2002 11:00 AM
To: dhcwg@ietf.org
Subject: [dhcwg] dhcpv6-24: randomization delay before invoking DHC

> 17.1.2. Transmission of Solicit Messages
> 
>    The first Solicit message from the client on the interface MUST
>    be delayed by a random amount of time between MIN_SOL_DELAY and
>    MAX_SOL_DELAY. In the case of a Solicit message transmitted when DHCP
>    is initiated by IPv6 Neighbor Discovery, the delay gives the amount
>    of time to wait after the ManagedFlag changes from FALSE to TRUE (see
>    section 5.5.3 of RFC2462).  This random delay desynchronizes clients
>    which start at the same time (for example, after a power outage).

Hmm. I see a couple of potential problems here. First, we have to be
careful about tying this to the change of the value from FALSE to
TRUE. Consider a misconfiguration, where there are two routers, and
one sends out RAs with the bit set to 0, the other router sets it to
1. This will result in a flip-flop with every RA. Do we really want to
restart DHC everytime that happens?

A better approach might be to say, if the managed bit changes from
FALSE to TRUE, *and* the client isn't currently already using DHC on
that interface, then start the process. note: if the flag changes from
true to false, does the client drop all its DHC-learned stuff and
issue a release?  There is no text that suggests this be done, and I
don't think we want to do this. 

Also, the delay should be tied to receipt of the first RA that says
use DHC. On a power outage, all the clients will get the first RA at
the same time. This is the issue where having a random delay would
seem to be most important.

Thomas

_______________________________________________
dhcwg mailing list
dhcwg@ietf.org
https://www1.ietf.org/mailman/listinfo/dhcwg