>From: "Crist J. Clark" <[EMAIL PROTECTED]>
>Reply-To: [EMAIL PROTECTED]
>To: Vadim Pushkin <[EMAIL PROTECTED]>
>CC: [EMAIL PROTECTED]
>Subject: Re: Need Help with Cisco sw VPN behind IpFilter/OpenBSD
>Date: Fri, 16 Aug 2002 09:30:16 -0700
>
>On Fri, Aug 16, 2002 at 01:28:38AM +0000, Vadim Pushkin wrote:
> > >From: "Crist J. Clark" <[EMAIL PROTECTED]>
> > >Reply-To: [EMAIL PROTECTED]
> > >
> > >[snip]
> > >
> > >> >> # 192.168.1.0 is my Unnatted/internal network.
> > >> >> #
> > >> >> pass in proto esp from vpn.XXX.net/32 to 192.168.1.0/24
> > >> >> pass out proto esp from 192.168.1.0/24 to vpn.XXX.net/32
> > >> >
> > >> >Find out if you are using ESP or 10000/udp.
> > >>
> > >> I am seeing alot of isakmp traffic getting blocked. I _believe_
> > >> that it is ESP.
> > >
> > >You seem to be very confused. ISAKMP rides over 500/udp. UDP is an IP
> > >protocol (protocol 17 to be exact). ESP, Encapsulated Security
> > >Payload, is a completely separate IP protocol (protocol 50).

The VPN that I am trying to connect to uses udp for authentication,
then esp for encrypted traffic.

> >
> > Thank you. What I am seeing is the following from tcpdump, but what
> > puzzles me is the fact that I sometimes see VPN.XXX.NET.IP, and some
> > times I see VPN.XXX.NET-ROUTER.IP, which are not even the same subnet.
> > See belows output sample.
> >
> > 21:05:38.838016 204.177.198.17.isakmp > VPN.XXX.NET.IP.isakmp:  isakmp 
>v1.0
> > exchange INFO encrypted
> >        cookie: 77ebe015bdc5ecce->2b4ea8a31d8bf87f msgid: fe44266a len: 
>84
> > 21:05:38.855278 VPN.XXX.NET.IP.isakmp > 204.177.198.17.isakmp:  isakmp 
>v1.0
> > exchange INFO encrypted
> >        cookie: 77ebe015bdc5ecce->2b4ea8a31d8bf87f msgid: e58d1012 len: 
>84
> > 21:05:49.853889 204.177.198.17.isakmp > VPN.XXX.NET.IP.isakmp:  isakmp 
>v1.0
> > exchange INFO encrypted
> >        cookie: 77ebe015bdc5ecce->2b4ea8a31d8bf87f msgid: afa02b26 len: 
>76
>
>Looks reasonable.
>
> > 21:05:49.854443 204.177.198.17.10346 > VPN.XXX.NET.IP.10000:  udp 1
> > 21:05:49.871418 VPN.XXX.NET-ROUTER.IP > 204.177.198.17: icmp: host
> > VPN.XXX.NET-ROUTER.IP unreachable - admin prohibited filter
>
>OK, it looks like a firewall at the remote site, VPN.XXX.NET-ROUTER.IP,
>is blocking the VPN packets (10000/udp). Run tcpdump with '-vvv' to
>print all the detail it can, especially to print the details about the
>header of the packet that caused the ICMP error message.

Further investigation reveals that VPN.XXX.NET-ROUTER.IP does indeed
block all icmp traffic, yet others are able to connect and their routers, 
D-Link, etc, do not even try to perform any icmp traffic.
Is there any reason why I am trying to use icmp? Is there any way to
prevent that?

>But it seems pretty clear the remote site is blocking the VPN. They
>will need to modify the policy on VPN.XXX.NET-ROUTER.IP to allow
>10000/udp from your network to go to VPN.XXX.NET.IP.

Yes, I too see this and agree. Thank you.

> > >> pass out quick on sis0 proto tcp from any to any flags S keep state
> > >> pass out quick on sis0 proto udp from any to any keep state
> > >> pass out quick on sis0 proto icmp from any to any keep state
> > >> pass out quick on sis0 proto esp from any to any keep state
> > >> #
> > >> pass in quick on sis1 proto tcp from any to 192.168.1.254 port = 22
> > >flags
> > >> S/SA
> > >> pass out quick on sis1 proto tcp from 192.168.1.254 port = 22 to any
> > >flags
> > >> S/SA
> > >
> > >A TCP SYN with a _source_ port of 22?
> >
> > ssh?
>
>ssh _listens_ on port 22. The client will send a SYN _to_ 22. Only the
>first of those two rules makes any sense.

I'll give it a shot. Again, thanks.

>
> > >> pass out quick on sis1 proto icmp all icmp-type 0
> > >> pass out quick on sis1 proto icmp all icmp-type 3
> > >> pass out quick on sis1 proto icmp all icmp-type 11
> > >> #
> > >> pass out quick on sis1 proto udp from 192.168.1.254 port = 53 to any
> > >> pass out quick on sis1 proto udp from 192.168.1.0 port = 53 to any
> > >
> > >Huh?
> >
> > Allows my internal machines to resolve against an external DNS server, 
>no?
>
>It allows the host at 192.168.1.254 to do UDP queries. It allows the
>host at 192.168.1.0 to do UDP queries. But I would guess you do not
>actually have a host at 192.168.1.0. The IP addresses 192.168.1.1
>through 192.168.1.253 cannot do UDP queries. Is that the intended
>policy?

No, I allow all hosts within the network to do udp queries, and it works as 
intended.

> > >> pass in quick on sis0 proto tcp from any to any port = 25 flags S/SA
> > >keep
> > >> state
> > >> pass in quick on sis0 proto tcp from any to any port = 29204 flags 
>S/SA
> > >keep state
> > >> pass in quick on sis0 proto tcp from any to any port = 22 flags S/SA
> > >keep state
> > >> block in log quick on sis0 proto tcp all flags S/SA
> > >> block out log quick on sis0 proto tcp all flags S/SA
> > >> block out on sis1 proto tcp all flags S/SA
> > >> block out on sis1 proto udp all
> > >> block out on sis1 proto icmp all
> > >
> > >You still seem to be missing some rules on sis1. I don't see how any
> > >of this works. I don't see how you can ever get a connection from your
> > >internal network out to the Internet.
> >
> > But I tell you it does :-)
>
>Well, I still don't see how anything gets out.

Dunno, try it for yourself, or maybe Mr. Reed can assist?

-vadim

Vadim (Ukranian Stallion) Pushkin


_________________________________________________________________
Send and receive Hotmail on your mobile device: http://mobile.msn.com

Reply via email to