On 11/23/10 2:23 AM, Julian Elischer wrote:
On 11/22/10 10:56 PM, kalin m wrote:
On 11/23/10 12:50 AM, Julian Elischer wrote:
On 11/22/10 6:14 PM, kalin m wrote:
On 11/22/10 8:43 PM, Julian Elischer wrote:
On 11/22/10 5:16 PM, kalin m wrote:
On 11/22/10 7:42 PM, Julian Elischer wrote:
On 11/22/10 3:52 PM, kalin m wrote:
thanks... got the ttyv8 off...
any idea about the temporary network loss?
not without more info...
it might be you machine or the switch it's attached to..
you may also have a clash with another machine somewhere with
the same IP address.
i forgot to mention one important thing... i can always ping the
machine from within the same network. so it seems that it's not
really the machine itself. the theory about the switch could be -
except all the other machines that are on the same network do not
have that problem. they are always pingable from the outside,
even when the machine in question is not.
and of course there is no other machine with the same ip on that
network. i did check...
so have you left tcpbump running on the machine looking at the
interface in question?
tcpdump -i XX0 -p icmp
this should tell you if it's the incoming or outgoing packets that
are getting lost.
i did similar stuff. but the thing is it's not only icmp. it's tcp
too. basically the machine is unreachable form the outside in any
given time. until i ssh into it from inside. http is not responding
either. and if i leave the dcpdump dumping - detached from the
terminal - it might run out of space....
but you didn't answer the question...
when icmp fails (ping) which packets are seen?
i don't know. i'm at home right now. it was off again. i had to go
ssh into it through another machine on the same network. but the
second i do that the ping from outside comes alive.
tomorrow at work will wait for this to happened and will run tcpdump
off the terminal of the machine itself. will report...
right now it just looks like this:
...............................................................................................
01:48:01.719224 IP machine.one.com > machine.two.com: ICMP echo
request, id 3386, seq 420, length 64
01:48:01.719250 IP machine.two.com > machine.one.com: ICMP echo
reply, id 3386, seq 420, length 64
01:48:02.719065 IP machine.one.com > machine.two.com: ICMP echo
request, id 3386, seq 421, length 64
01:48:02.719090 IP machine.two.com > machine.one.com: ICMP echo
reply, id 3386, seq 421, length 64
01:48:03.724056 IP machine.one.com > machine.two.com: ICMP echo
request, id 3386, seq 422, length 64
01:48:03.724082 IP machine.two.com > machine.one.com: ICMP echo
reply, id 3386, seq 422, length 64
just write it to a file so you cna see what was going on when it was
"asleep"
ok. just afraid it will run out of space while waiting for this to
happen again.... it's running now...
I suspect a MAC setting on the switch or something.
does the 'ssh' come in on the same interface as the ping?
yes.
please check the netmasks everywhere on the router, on the machine
and on other local machines.
well... "the router" is a actually out of my reach. it's a fios set up.
the netmask for all the machines is 255.255.255.0 so the actual router
is somewhere in the building. i guess. we just got 13 ips assigned to
the office but it's not like a subnet set up. there is a box in the
office before the machines but my guess is it's not really a router...
i do not have access to it either. does this make any sense?
thanks again for sticking with this...
..........................................................................................
where machine.two.com is the one that disappears....
thanks...
thanks...
thanks....
On 11/22/10 6:25 PM, Julian Elischer wrote:
On 11/22/10 2:47 PM, kalin m wrote:
hi all..
recently i had to change the static ip on a machine here at
work. nothing wrong with that. what's happening with this new
ip (different network) is that for short periods of time
sometimes the machine is unpingable. until i try to login
ssh. as soon as i do that the ping and the rest of network
activity resumes - http.
that didn't use to happen on the previous network. the only
change was the network information. also a bunch of other
machines changed at the same time and none of the others
present the same situation.
the only entry in the log is:
init: can't exec getty '/usr/local/bin/xdm' for port
/dev/ttyv8: No such file or directory
delete or turn off the line in /etc/ttys for ttyv8
(if somebody would explain what this is, that would help too.
since xdm is not really in use and i can't find anything that
would be calling xdm.)
not sure if this has any importance but the card itself is
BCM5750A1 NetXtreme Gigabit Ethernet PCI Express according to
pciconf.
the machine is 7.2 release.
where do i look?
thanks...
_______________________________________________
freebsd-net@freebsd.org mailing list
http://lists.freebsd.org/mailman/listinfo/freebsd-net
To unsubscribe, send any mail to
"freebsd-net-unsubscr...@freebsd.org"
_______________________________________________
freebsd-net@freebsd.org mailing list
http://lists.freebsd.org/mailman/listinfo/freebsd-net
To unsubscribe, send any mail to "freebsd-net-unsubscr...@freebsd.org"
