Re: WG ACTION: 1 week to discuss [LL37] Aggressive Time-outs
"Philip Nye" <[email protected]> Thu, 18 Dec 2003 10:26:26 -0000
| Newsgroups | gmane.ietf.zeroconf |
|---|---|
| Organization | Engineering Arts |
| Message-ID | <033b01c3c551$64eb7040$131010ac@aldebaran> |
> From: "Bill Rees" <[email protected]> > > As I understand this thread, the latency comes from waiting for a > spanning tree routine to run which to me implies a wan, no? It seems that spanning tree can take quite a bit longer so the difference between 2 and 8 seconds is little help. > So a bit > off kilter here, what is the likelihood that the llv4 situation occurs > in such a congested, logically disparate locations? Is it likely that > linklocal will show up in such an environment? Given all the discussion > surrounding moving off of the llv4 address as quickly as possible, even > in a congested environment what is the likelihood that the number of > llv4 devices will exist on the network at the same time isd 2 or more? The answer to this was discussed a long time ago. If the network has a DHCP server, which refuses to hand you an address or if it is primarily a statically configured network, you can still go ahead and configure a LL address and use it to whatever extent it works (e.g. to talk to a printer or another ad-hoc user). This could easily apply on a large multiply bridged network. Philip