[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
Re: MacGate Revival
On Friday, May 17, 2013 6:00:55 AM UTC-5, Miles Lott wrote:
> On Friday, May 17, 2013 5:59:46 AM UTC-5, Miles Lott wrote:
>
> > On Thursday, May 16, 2013 6:53:10 PM UTC-5, Steven Hirsch wrote:
>
> >
>
> > > On 05/16/2013 07:48 PM, Jonno Downes wrote:
>
> >
>
> > >
>
> >
>
> > > > On Friday, May 17, 2013 7:36:15 AM UTC+10, Miles Lott wrote:
>
> >
>
> > >
>
> >
>
> > > >> Yes, telnet is running on the a2server install specifically for this ;)
>
> >
>
> > >
>
> >
>
> > > >> I was able to connect to localhost and then from another linux vm, my
>
> >
>
> > >
>
> >
>
> > > >> primary file server. iptables is disabled (policy ACCEPT with no further
>
> >
>
> > >
>
> >
>
> > > >> rules).
>
> >
>
> > >
>
> >
>
> > > >>
>
> >
>
> > >
>
> >
>
> > > > Can you use wireshark to confirm whether any of the telnet traffic is
>
> >
>
> > >
>
> >
>
> > > > leaving the PN848?
>
> >
>
> > >
>
> >
>
> > > >
>
> >
>
> > >
>
> >
>
> > >
>
> >
>
> > >
>
> >
>
> > > Excellent idea! It wouldn't be the first time a LocalTalk <--> Ethernet
>
> >
>
> > >
>
> >
>
> > > bridge has caused problems. Geoff has been on ground zero for all the war
>
> >
>
> > >
>
> >
>
> > > stories :-).
>
> >
>
> >
>
> >
>
> > tcpdump shows quite a bit of AT traffic, including:
>
> >
>
> >
>
> >
>
> > 06:55:54.078248 AT 1.146.72 > 1.105.72: at-IP 40
>
> >
>
> > 06:55:57.126493 AT 1.146.72 > 1.105.72: at-IP 40
>
> >
>
> >
>
> >
>
> > ...which should be the MacIP packets. The issue seems to be with the ability to actually see this traffic in MacGated. If I set the socket to listen to ETH_P_ALL then it sees plenty that it cannot decode. I added some debugging to show them but I should spend more time with that. For the socket I have tried:
>
> >
>
> >
>
> >
>
> > // SockLT = socket(AF_PACKET, SOCK_RAW, htons(ETH_P_LOCALTALK));
>
> >
>
> > // SockLT = socket(AF_APPLETALK, SOCK_DGRAM, htons(ETH_P_ALL));
>
> >
>
> > // SockLT = socket(AF_PACKET, SOCK_DGRAM, htons(ETH_P_LOCALTALK));
>
> >
>
> >
>
> >
>
> > also htons(ETH_P_ATALK), but only the ETH_P_ALL shows anything so far. So, only the most critical parts are left to fix ;)
>
>
>
> Note that I am also running netatalk since that seems to be critical for nbp registration, etc. That part works fine.
I have sent my current work to Geoff for his evaluation.