cpghost wrote:
Could you provide tcpdump output for connection tries from your
Ethernet interface? Use "-pes 0" options please.
Will do; but I'll first have to wait 24h from now to get a
forcibly disconnected session (I've just had to restart ppp
again).

All right, I've got a good tcpdump now:

# tcpdump -i sis0 -n -pes 0
tcpdump: verbose output suppressed, use -v or -vv for full protocol
decode listening on sis0, link-type EN10MB (Ethernet), capture size
65535 bytes

17:06:08.400881 00:00:24:c2:45:74 > ff:ff:ff:ff:ff:ff,
  ethertype PPPoE D (0x8863), length 32: PPPoE PADI [Host-Uniq
  0xC0C263C1] [Service-Name]

17:06:08.400891 00:00:24:c2:45:74 > ff:ff:ff:ff:ff:ff, ethertype PPPoE
  D (0x8863), length 32: PPPoE PADI [Host-Uniq 0xC0ED45C1]
  [Service-Name]

17:06:08.400898 00:00:24:c2:45:74 > ff:ff:ff:ff:ff:ff, ethertype PPPoE
  D (0x8863), length 32: PPPoE > PADI [Host-Uniq 0x40C263C1]
  [Service-Name]

Your host generates huge number of simultaneous session requests. Looking on different Host-Uniq values it should be different ng_pppoe sessions/hooks as Host-Uniq is actually pointer to the hook/session internal data and it stays persistent for all packets of the same session. So it looks that something creates those many hooks.

I think such behaviour can take place if ppp daemon for some reason don't waits for reply but closes session immediately after sending connect request. If it so it also explains original "no matching session" errors as for the answer received time session/hook can already be destroyed.

Provide please your ppp configuration files and part of detailed log file (set log All) describing connection attempts.

PS: I am not ppp daemon expert, but quick code look shows that connection waiting can be configured with 'set cd ...' command. Setting 'set cd 0' gives results close to yours. Have you something like that in your configuration?

--
Alexander Motin
_______________________________________________
freebsd-stable@freebsd.org mailing list
http://lists.freebsd.org/mailman/listinfo/freebsd-stable
To unsubscribe, send any mail to "[EMAIL PROTECTED]"

Reply via email to