Re: [pfSense Support] Is there any reason I can't Remote desktop through an ipsec tunnel?

2009-03-27 Thread Adam Armstrong

Marty Nelson wrote:


I have an IPSec tunnel connecting my network to one of our customer 
sites, and while I can ping a computer on their network I am unable to 
remote desktop to. Currently all of our customer tunnels are setup to 
terminate in our DMZ to limit access back into our network. I have a 
second firewall (monowall) in our DMZ that then routes all traffic out 
through the tunnel. I’ve drawn a rudimentary layout of how it’s setup 
(see below).


I have the IPsec rules to pass all traffic, and currently I have it 
setup to log all traffic as well. What’s strange is that when I 
attempt to remote desktop to it, I see no traffic relating to that at 
all. Nothing passing, nothing getting blocked. Like I said, I can ping 
the box just fine (and it shows up in the log), but I am unable to 
remote desktop to it and I don’t see anything getting blocked, or passed.


Hopefully this made sense. If it’s unclear, please let me know and 
I’ll try my best to clear it up.


LAN (192.168)---[pfSenseFW]---DMZ (10.100)---[monowall]---[ipsec 
tunnel to cust site]---Cust site


I would say that it's almost certainly MTU-related. RDP always seems to 
be the first thing hit by a failure of the pmtud mechanism to work.


The IPSEC tunnel will be reducing your MTU, and when the RDP server 
tries to send out a packet it'll get dropped. Try reducing the MTU of 
the interface of the server?


This usually manifests itself by the login screen background appearing 
(presumably because it fits into  1492 bytes), but then nothing more. 
Doesn't sound exactly like what you're seeing, but RDP + IPSEC issues 
are usually MTU-related IME.


adam.

-
To unsubscribe, e-mail: support-unsubscr...@pfsense.com
For additional commands, e-mail: support-h...@pfsense.com

Commercial support available - https://portal.pfsense.org



Re: [pfSense Support] BGP status

2008-02-11 Thread Adam Armstrong

Royce Mitchell III wrote:

Adam Armstrong wrote:
Carp is unnecessary when using BGP, as the provider sees routes into 
your network via the individual devices and both devices see routes out.


You wouldn't want to run BGP from a CARP IP anyawys, as it would 
result in BGP flapping when the CARP switched.


adam.
Okay, please forgive my ignorance, but if you have two redundant 
routers servicing your BGP, how will they decide who is going to 
handle a packet without some sort of CARP/VRRP communication between 
them?
There are a number of mechanisms for doing this, generally you'll set 
the localpref high for prefixes coming from the peer you want to use, 
and set the MED low for prefixes being announced to that peer, that way 
your peer will send traffic to you on the correct link (lowest MED wins) 
and you'll send traffic out on the correct link (highest localpref wins).


However, if you're doing BGP solely to get redundant connectivity to the 
same ISP you should look again at CARP and ask what your ISP can do by 
way of HSRP/VRRP to present a single IP to you from two of their 
devices. VRRP/CARP/HSRP is generally a far better solution for that due 
to the slowlness of BGP convergence.


adam.

-
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]



Re: [pfSense Support] BGP status

2008-02-11 Thread Adam Armstrong

Paul M wrote:

Royce Mitchell III wrote:
  

Is the BGP package for pfsense available, yet?

Also, does it play nice with CARP, or is CARP even necessary when you
have BGP?



I think CARP is a very different thing - BGP is a way of having multiple
circuits to different ISPs to get resilience internet connectivity. CARP
is a way of having two devices share an IP.

Or am I missing some clever use of BGP and CARP?
  
Carp is unnecessary when using BGP, as the provider sees routes into 
your network via the individual devices and both devices see routes out.


You wouldn't want to run BGP from a CARP IP anyawys, as it would result 
in BGP flapping when the CARP switched.


adam.

-
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]



Re: [pfSense Support] Spanning tree support

2008-02-07 Thread Adam Armstrong

Gary Buckmaster wrote:

Chris Bagnall wrote:

Greetings list,

Does anyone know if pfSense includes support for failover between two 
LAN interfaces?


For example, one can provide high availability using CARP to create a 
virtual router IP failing over between 2 pfSense boxes, but that's 
not going to solve the problem of a switch dying. It'd be useful to 
be able to connect 2 interfaces from each box to the LAN (one to each 
switch), then configure them using spanning tree protocol (or one of 
the derivatives).


If it's not currently included, are there plans to do so, and/or what 
sort of financial incentive would encourage development on that 
front? :-)


Regards,

Chris
  

Chris,

There's been a call on the site for awhile for some hardware that 
support STP.  I don't know if that call is still valid or if they got 
hardware in.  I suspect that you'd want to consider a bounty project 
or get in touch with BSDPerimeter and put together a formal quote.  I 
hope you choose to pursue this, it'd be a nice feature to have.
You don't need spanning tree support on the router to accomplish this. 
You just need NIC 'teaming' support in the OS.


Linux supports this in a variety of modes, for example, using a single 
MAC address across two ports but only transmitting on one, or using 
standards-based link aggregation  to allow the bandwidth of both 
connections to be used (you could use this with a stacking switch such 
as a 3750 to also get resilience).


I would guess that FreeBSD also has support for this somewhere, it would 
just be a case of building it into the back-end and web interface.


STP is, in my opinion, a brain-dead way of accomplishing this. STP 
should be eliminated from any well-designed modern network wherever 
possible!


Thanks,
adam.

-
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]



Re: [pfSense Support] Issue with stalling on static route

2007-09-26 Thread Adam Armstrong

jamespev wrote:


Hello all!  I am having a major issue that I'm hoping you can shed 
light on.  We recently added an MPLS link from our location to our 
other company offices (replacing a pfsense VPN tunnel that was working 
great) and am now having issues across it.  The MPLS is hooked to a 
cisco router sitting behind our pfsense firewall, and I setup a static 
route on pfsense over to it for the appropriate subnet.  This seemed 
to work fine, but after using it a bit it seems that traffic is 
getting stalled somewhere.  If I setup a static route on my desktop 
machine (client machine on network) to the cisco (for the appropriate 
subnet) everything works perfectly.  So it seems something is 
happening on the pfsense machine.  Shorter transactions seem to be 
fine, pinging always works.  Outlook however is very unhappy 
(consequently so are the users...).  In general it seems that TCP 
services being effected most.
I did a packet capture with and without the static route on my 
client machine.  With all the traffic going through the pfsense there 
were a lot of TCP retransmissions happening.
Could this be an issue with pfsense's packet scrubbing?  There is 
nothing in the firewall logs to indicate that anything is being 
blocked.  I am using 1.2RC2.
If anyone has any ideas I would be very appreciative.  I think the 
users are starting to gather torches and pitchforks...


James

You haven't specified what MPLS-based service(s) you're taking!

First point of call for all MPLS-related issues : have you made sure you 
can pass full 1500-byte frames across the circuit?


adam.





-
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]



Re: [pfSense Support] British Telecom and pfSense

2007-08-14 Thread Adam Armstrong

Siju George wrote:

Dear People in U.K,

I had a little taste of the British Telecom ( BT ) ISP a while back.

It seems that they have a device ( modem ) that has to be used to give
IP address by dhcp to all systems.

And even if you want to give Static IPs ( routable on the Internet )
assigned to you by BT on any of your servers you should first assign
the server an IP in the local range through dhcp and then get into the
BT device's web interface and assign the Static IP through that
interface to the appropriate server.

How do the British people use pfSense as the firewall while using BT
as their ISP?
I asked the sysadmin there and he said they just use the firewall on
the BT device (modem) and that the device cannot be done away with or
configured to allow us to define static IPs on the interface of our
Servers using the config files on the servers Operating System.

Is this true?
And what are the other ISPs other than BT in U.K especially in the Bolton Area?
  
Almost all ADSL ISPs in the UK are national, that is they share the BT 
ADSL network so get almost 100% national reach.


You need to find an ISP which will allow you to use any routing hardware 
you like, and configure it how you like, possibly giving you a subnet 
for use with the device. I wasn't aware of BT having any draconian 
limitations, unless you go for their BT HomeHub service, which of 
course, needs to use the home hub!


Almost any 'decent' UK ADSL ISP should suffice, and of course the 
company I work for provides ADSL too :)


There have been a large number of  highly 'commoditised' ADSL providers 
eating up the market at the moment, such as Orange, Sky and Tiscali. If 
this is for a business application where you need some IPs, you should 
look elsewhere than the 'cheap' providers here.


There are a few options for how you'd get ADSL connectivity to the 
pfsense sanely, you could use a Draytek Vigor 100 to pass a PPPoE 
connection to the pfSense, or you could use something like a Zoom X3 to 
do 'half bridge' and present the same connectivity to you on the 
ethernet to the pfSense as it's presented by the remote ADSL LNS.


Hope that helps!

adam.



-
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]



Re: [pfSense Support] performance on a PE860

2007-08-13 Thread Adam Armstrong

jamona perez wrote:
Thanks for the tip, I'm not too sure about this stp stuff, because I 
always think twice before doing that kind of stuff, I've had my share 
of network loops not being always well-handled by switch hardware. On 
the other hand I've read from m0n0wall's forum that it is feasible.

so if it's the way to go, I'll go.
Last, (I don't wan't to start flame war, please), as all I want to do 
is transparent FW, maybe I should go for m0n0wall instead of pfsense. 
The drawback of monowall being that it won't support smp, thus making 
me stick to a single celeron 3.33 Ghz, and running freebsd 4.2 (will 
the double-port intel pcie card be supported ?).


The Intel nic should be easily supported, though others may know id 
there have been lockups or other nasties with the old freebsd version.


There is a benefit with monowall, in that it's quite a bit faster.

The major downside is that you can't synchronise rules across the two 
firewalls, as you could by using pfsync with pfSense (note, i don't mean 
using carp, just rule synchronisation!)


If you do try it, let me know how it works, because i may be interested 
in doing it too!


regards,
adam.

regards




 Date: Fri, 10 Aug 2007 23:49:13 +0100
 From: [EMAIL PROTECTED]
 To: support@pfsense.com
 Subject: Re: [pfSense Support] performance on a PE860

 jamona perez wrote:
  Okay, I did not realize that, this is really helpful info. Thinking
  about it for 2 minutes I just realized that a in bridge mode, the WAN
  does not really have an IP address, does it ? so carp has no IP
  failover to do whatsoever.
  Please Correct me if I'm wrong.
  So if the best I can do is having a spare box standing by to get
  fired up if the other goes down ,it's what i'm going to do. But if 
you

  can think of any mecanism (similar to linux heartbeat) that can sit
  here waiting for the other side to fail, then take the appropriate
  measure (read configurable like starting the proper services) to
  ensure high-availability of such a system, I'll be more that glad to
  hear about it.
 
 If pfSense will allow you to pass STP frames across it, you could just
 put two pfSense boxes in parallel like so

 EXTERNAL SWITCH
 FA0/1 FA0/2
 | |
 | |
 FW1--SYNC-- FW2
 | |
 | |
 FA0/1 FA0/2
 INTERNAL SWITCH



 Assuming that STP will pass the packets, you should have no issues in
 this configuration. STP will put the ports of FA0/2 into blocking mode,
 and no traffic will pass unless traffic stops flowing across FA0/1 
(yes,

 i have just realised that you were probably meaning gig interfaces, but
 i did the diagram already :P)

 Someone else here will probably better know wether or not you can pass
 STP across pfSense correctly...

 You might also want to use two more interfaces for management? (don't
 give the firewalls IPs on the bridge, so that FW2 is still accessable
 when the links are blocking!)

 adam.

 -
 To unsubscribe, e-mail: [EMAIL PROTECTED]
 For additional commands, e-mail: [EMAIL PROTECTED]




Besoin d'un e-mail ? Créez gratuitement un compte Windows Live Hotmail 
et bénéficiez de 2 Go de stockage ! Windows Live Hotmail 
http://www.windowslive.fr/hotmail/default.asp



-
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]



Re: [pfSense Support] performance on a PE860

2007-08-10 Thread Adam Armstrong

jamona perez wrote:
Okay, I did not realize that, this is really helpful info. Thinking 
about it for 2 minutes I just realized that a in bridge mode, the WAN 
does not really have an IP address, does it ? so carp has no IP 
failover to do whatsoever.

Please Correct me if I'm wrong.
So if the best I can do is having a spare box standing by to get 
fired up if the other goes down ,it's what i'm going to do. But if you 
can think of any mecanism (similar to linux heartbeat) that can sit 
here waiting for the other side to fail, then take the appropriate 
measure (read configurable like starting the proper services) to 
ensure high-availability of such a system, I'll be more that glad to 
hear about it.
 
If pfSense will allow you to pass STP frames across it, you could just 
put two pfSense  boxes in parallel like so


EXTERNAL SWITCH
FA0/1FA0/2
   |  |
   |  |
FW1--SYNC-- FW2
   |  |
   |  |
FA0/1   FA0/2
INTERNAL SWITCH



Assuming that STP will pass the packets, you should have no issues in 
this configuration. STP will put the ports of FA0/2 into blocking mode, 
and no traffic will pass unless traffic stops flowing across FA0/1 (yes, 
i have just realised that you were probably meaning gig interfaces, but 
i did the diagram already :P)


Someone else here will probably better know wether or not you can pass 
STP across pfSense correctly...


You might also want to use two more interfaces for management? (don't 
give the firewalls IPs on the bridge, so that FW2 is still accessable 
when the links are blocking!)


adam.

-
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]



Re: [pfSense Support] Failover from PC to WRAP

2007-06-28 Thread Adam Armstrong


Is it possible to have a 'hot standby' on a WRAP if the firewall 
is a PC (we are not running any packages).  Don't consider performance.



There's no reason why not!

adam.

-
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]



Re: [pfSense Support] pfSense Firewall Logs: no ports listed !?

2007-06-16 Thread Adam Armstrong
224.0.0.2 is the all routers multicast address, and any traffic to it 
is probably router discovery or something similar.


adam.

That looks more like a protocol decode issue to me.  224.0.0.2 is a
multicast address, I wouldn't be surprised if that really wasn't UDP.
Can you show an example of a TCP log entry w/out ports, or something
to a non-multicast address?  Thanks

--Bill

On 6/16/07, Heiko Garbe [EMAIL PROTECTED] wrote:

Hello,
here is a screenshot. I think he means the firewall logs in the gui
Greetings
heiko

Chris Buechler schrieb:
 On Fri, 2007-06-15 at 18:01 +0200, Fuchs, Martin wrote:

 Hi !

 In the firewall logs always was shown blocked traffic with the ports
 that were used...

 Now with the 6-6 snapshot it does not display the ports anymore 
... !?


 It's a little confusing and seems tob e a bit silly / senseless 
not to

 display the ports !?



 Can you post a screenshot? Not sure exactly what you mean, I haven't
 seen or heard of any issues.



 -
 To unsubscribe, e-mail: [EMAIL PROTECTED]
 For additional commands, e-mail: [EMAIL PROTECTED]




-
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]




-
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]




-
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]



RE: [pfSense Support] SNMP

2007-02-21 Thread Adam Armstrong
 On 2/21/07, Andrew Kemp [EMAIL PROTECTED] wrote:
  any plans to enable additional SNMP'able items like cpu usage,
 memory
  usage, and disk usage? i know m0n0wall allowed me to graph a few more
  values in cacti than pfsense does.
 
 As soon as someone adds the support to bsnmpd, sure.
 

Are there any reasons why pfSense doesn't use net-snmpd?

Adam.


-
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]



[pfSense Support] CARP address bug

2007-02-16 Thread Adam Armstrong
Hi,

 

If I try to edit a CARP virtual IP, it tells me that the VHID is already in
use and won't let me save the changes. The result is that every time I
change a setting on one of the CARP virtual IPs, I have to change the VHID
on every device. Very annoying!

 

Is this intended behaviour, or a horrid bug?

 

Thanks,

Adam.



RE: [pfSense Support] pfSense SNMP identification

2007-01-31 Thread Adam Armstrong
 On 1/25/07, Adam Armstrong [EMAIL PROTECTED] wrote:
 [snip]
  Except, Still no suggestions on how to identify a pfSense needle in a
  haystack of FreeBSD servers. :\
 
 Just let us know what you want us to do and we will do it.
 

Manuel has made a modification to m0n0wall to give out 'correct' snmp
sysDescr values : 

[EMAIL PROTECTED]:~# snmpwalk -v2c -c xxx w3z.alderwasley.org
SNMPv2-MIB::sysDescr.0 = STRING: m0n0wall w3z.alderwasley.org 1.23b3
net45xx FreeBSD 4.11-RELEASE-p26 i386

Based on the format : 

$os $host $version $platform $base $base_version $base_hardware

Could you guys make this modification to pfsense too?

Thanks,
Adam.


-
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]



[pfSense Support] Embedded bootup

2005-11-23 Thread Adam Armstrong
Hi,

I have a Lex box (the CV860A), but I can't get it to boot any version of
pfSense.

I'm trying to use the embedded image, but every one I've tried over the past
few weeks has hung at right after 'Loading /boot/defaults/loader.conf'.

The little spinning line spins a few times and then just stops. I've tried
changing the bios settings I thought would make a difference, but to no
avail.

I'm currently trying 0.94.10, but I've tried a lot of other images and every
one of them locks at the same point!

Kind Regards,
Adam.


-
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]



[pfSense Support] SNMP / Updating

2005-11-15 Thread Adam Armstrong
Hi All,

My pfsense devices won't do auto update, are there any known issues with
0.90?

Also, the snmp daemon on my 0.90 devices are reporting that their interfaces
are all down! I'm using carp, will this be causing it?

Thanks,
Adam.


-
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]