dccd vs. EADDRNOTAVAIL

Vernon Schryver vjs@calcite.rhyolite.com
Fri Oct 6 15:15:47 UTC 2006


z.dcc-servers.net is now running on an amd64 CPU with FreeBSD 6.1.  A
few dozen times per day, dccd whines that the sendto() system call
used to answer a client request resulted in EADDRNOTAVAIL.

Poking through /usr/src/sys suggests to me that EADDRNOTAVAIL cannot
happen from the normal UDP/IP code, at least not unless dccd is sending
to a multicast address.  (It's not).  I did find a place or two that
look to me like IPFW kernel hooks that might but shouldn't result in
EADDRNOTAVAIL.

I hate to just ignore errors, but I also hate logs filled with useless
noise that hides the stuff that one should watch.  Has anyone else see
much of EADDRNOTAVAIL?  Or know what it's about?


Vernon Schryver    vjs@rhyolite.com



More information about the DCC mailing list

Contact vjs@rhyolite.com by mail or use the form.