[CUWiN-Dev] Re: [CUWiN-Commits] r3602 - cuw/trunk/src/hsls

David Young dyoung at pobox.com
Fri Oct 21 16:18:14 CDT 2005


On Fri, Oct 21, 2005 at 10:15:27AM -0500, Bill Comisky wrote:
> On Fri, 21 Oct 2005, dyoung at cuw.ojctech.com wrote:
> 
> >Author: dyoung
> >Date: 2005-10-21 00:17:02 -0500 (Fri, 21 Oct 2005)
> >New Revision: 3602
> >
> >Modified:
> >  cuw/trunk/src/hsls/hsls_router.c
> >Log:
> >Fix a serious regression in hslsd: now we count the number of
> >address families in a Hello *for real*.  hslsd wasn't forming
> >adjacencies because of this.
> 
> I'm not sure what this means; what would the routing table look like if it 
> was not "forming adjacencies"?  I have problems with out testbed this 
> week, maybe this was one of the issues.  I was seeing only 169.254 entries 
> for other nodes from the gateway no 10.0 entries.

Sounds like you were affected by the bug.  Anyone would be, if they ran
code from a couple days ago.

When I say a node Y forms an adjacency, I mean Y extracts from its
neighboring router's hello the information that neighbor X exists,
creates an hsls_peer object to keep state for that neighbor, and puts
a linkstate X->Y into its routing table.  The bug I fixed made every
hello packet look like it was empty, so a node would not extract any
information at all from a hello packet.

Dave

-- 
David Young             OJC Technologies
dyoung at ojctech.com      Urbana, IL * (217) 278-3933


More information about the CU-Wireless-Dev mailing list