Re: CARP & bridge

2009-05-01 Thread Nikos Vassiliadis

Hi,

Sebastiaan van Erk wrote:
So I don't really *NEED* the CARP ip address over the bridge (the static 
arp works, so I have a working solution, albeit an ugly one; an ARP 
request generates a reply from every member of the redundant cluster).


Just a guess, you could try adding the VIP/32 to the tap interface,
instead of the static arp thing. Don't know if it will work, it is
just a guess, which looks - to me - like a cleaner configuration.
At least it's rc.conf friendly.

Just my 0.2 euros, Nikos
___
freebsd-questions@freebsd.org mailing list
http://lists.freebsd.org/mailman/listinfo/freebsd-questions
To unsubscribe, send any mail to "freebsd-questions-unsubscr...@freebsd.org"


Re: CARP & bridge

2009-05-01 Thread Sebastiaan van Erk

Hi,

Nikos Vassiliadis wrote:

Sebastiaan van Erk wrote:


Thanks for the suggestion. I tried it, but unfortunately the carp 
device never leaves the INIT state when I put the ip on the bridge. 
:-( I did find some similar problem here:


http://www.freebsd.org/cgi/query-pr.cgi?pr=125816


I just noticed that. On -CURRENT carp tells you that's
not supported:
bridge0: carp is not supported for this interface type

OTOH why do you even have to use the VIP from the remote
side of the bridge?

The only reason I can think of, for doing  such a thing,
is to get *all* traffic from the remote location through
a "single" redundant router, the one with the VIP. Is this
the case?


It is indeed a "single" redundant router, though the traffic from the 
other side of the bridge (the OpenVPN clients) generally don't need to 
be routed redudantantly. The OpenVPN clients use OpenVPN's redundancy 
(multiple "remote xxx.xxx.xxx.xxx" lines), and thus use the 
non-redundant IP address of the OpenVPN client they're connected to as 
gateway (which is fine, because if the server dies OpenVPN connects to a 
different server anyway)...


So I don't really *NEED* the CARP ip address over the bridge (the static 
arp works, so I have a working solution, albeit an ugly one; an ARP 
request generates a reply from every member of the redundant cluster).


I guess it's just not a supported configuration yet and it's not my 
stupidity (in this case anyway ;-)) that's the problem.



Nikos


Regards,
Sebastiaan


smime.p7s
Description: S/MIME Cryptographic Signature


Re: CARP & bridge

2009-04-30 Thread Nikos Vassiliadis

Sebastiaan van Erk wrote:


Thanks for the suggestion. I tried it, but unfortunately the carp device 
never leaves the INIT state when I put the ip on the bridge. :-( I did 
find some similar problem here:


http://www.freebsd.org/cgi/query-pr.cgi?pr=125816


I just noticed that. On -CURRENT carp tells you that's
not supported:
bridge0: carp is not supported for this interface type

OTOH why do you even have to use the VIP from the remote
side of the bridge?

The only reason I can think of, for doing  such a thing,
is to get *all* traffic from the remote location through
a "single" redundant router, the one with the VIP. Is this
the case?

Nikos
___
freebsd-questions@freebsd.org mailing list
http://lists.freebsd.org/mailman/listinfo/freebsd-questions
To unsubscribe, send any mail to "freebsd-questions-unsubscr...@freebsd.org"


Re: CARP & bridge

2009-04-29 Thread Sebastiaan van Erk

Hi,

Nikos Vassiliadis wrote:

Sebastiaan van Erk wrote:

Julien Cigar wrote:


Maybe you've to do ARP Proxy on one side ? Try to add an ARP entry in
the ARP table with arp (arp -s 1.2.3.4 MAC foo) ..


Thanks for the suggestion.

Ok, static arp works: that is, if I take the carp1 mac address and add 
it to the arp table using:


 arp -s 10.0.80.74 00:00:5e:00:01:02 pub

The ping starts to work. I'm still a bit confused why I have to do 
this though, because I can ping the non-shared IP 10.0.80.77 from the 
VPN client (via tap0) without any static arp, and I can ping the 
shared VIP (10.0.80.74) from clients on the physical network (em1) as 
well without any static arp. It's only when the ping it has to cross 
the bridge that it's an issue.


Does it make any difference if you set the IP address on the bridge0
iface and not on the physical one?

I recall that the recommended setup is to use IP addresses on
the bridge interface and leave the members of the bridge IPless.

Nikos


Thanks for the suggestion. I tried it, but unfortunately the carp device 
never leaves the INIT state when I put the ip on the bridge. :-( I did 
find some similar problem here:


http://www.freebsd.org/cgi/query-pr.cgi?pr=125816

Regards,
Sebastiaan


smime.p7s
Description: S/MIME Cryptographic Signature


Re: CARP & bridge

2009-04-29 Thread Nikos Vassiliadis

Sebastiaan van Erk wrote:

Hi,

Julien Cigar wrote:

On Wed, 2009-04-29 at 11:37 +0200, Sebastiaan van Erk wrote:

Hi,

I have a bridged OpenVPN setup where the OpenVPN tap0 driver is 
bridged (via bridge0) to the physical em1 interface, which has a VIP 
via a carp1 interface:


em1: flags=8943 
metric 0 mtu 1500

options=98
ether 00:0c:29:61:2a:55
inet 10.0.80.77 netmask 0xff00 broadcast 10.0.80.255
media: Ethernet autoselect (1000baseTX )
status: active
bridge0: flags=8843 metric 0 
mtu 1500

ether 9a:6a:9f:b2:65:da
id 00:00:00:00:00:00 priority 32768 hellotime 2 fwddelay 15
maxage 20 holdcnt 6 proto rstp maxaddr 100 timeout 1200
root id 00:00:00:00:00:00 priority 32768 ifcost 0 port 0
member: tap0 flags=143
ifmaxaddr 0 port 11 priority 128 path cost 200
member: em1 flags=143
ifmaxaddr 0 port 2 priority 128 path cost 2
tap0: flags=8943 
metric 0 mtu 1500

ether 00:bd:48:03:00:00
Opened by PID 24616
carp1: flags=49 metric 0 mtu 1500
inet 10.0.80.74 netmask 0xff00
carp: MASTER vhid 2 advbase 1 advskew 0


The problem I have is that when I ping the VIP from a VPN client (on 
tap0), the server receives arp requests for the VIP on tap0, but it 
does not respond to them:


# tcpdump -i tap0 -ln
11:29:13.637048 arp who-has 10.0.80.74 tell 10.0.80.6

Is there any way to get the server to respond to arp requests on tap0 
for the VIP?




Maybe you've to do ARP Proxy on one side ? Try to add an ARP entry in
the ARP table with arp (arp -s 1.2.3.4 MAC foo) ..


Thanks for the suggestion.

Ok, static arp works: that is, if I take the carp1 mac address and add 
it to the arp table using:


 arp -s 10.0.80.74 00:00:5e:00:01:02 pub

The ping starts to work. I'm still a bit confused why I have to do this 
though, because I can ping the non-shared IP 10.0.80.77 from the VPN 
client (via tap0) without any static arp, and I can ping the shared VIP 
(10.0.80.74) from clients on the physical network (em1) as well without 
any static arp. It's only when the ping it has to cross the bridge that 
it's an issue.


Does it make any difference if you set the IP address on the bridge0
iface and not on the physical one?

I recall that the recommended setup is to use IP addresses on
the bridge interface and leave the members of the bridge IPless.

Nikos
___
freebsd-questions@freebsd.org mailing list
http://lists.freebsd.org/mailman/listinfo/freebsd-questions
To unsubscribe, send any mail to "freebsd-questions-unsubscr...@freebsd.org"


Re: CARP & bridge

2009-04-29 Thread Julien Cigar
On Wed, 2009-04-29 at 11:37 +0200, Sebastiaan van Erk wrote:
> Hi,
> 
> I have a bridged OpenVPN setup where the OpenVPN tap0 driver is bridged 
> (via bridge0) to the physical em1 interface, which has a VIP via a carp1 
> interface:
> 
> em1: flags=8943 metric 0 
> mtu 1500
>   options=98
>   ether 00:0c:29:61:2a:55
>   inet 10.0.80.77 netmask 0xff00 broadcast 10.0.80.255
>   media: Ethernet autoselect (1000baseTX )
>   status: active
> bridge0: flags=8843 metric 0 mtu 
> 1500
>   ether 9a:6a:9f:b2:65:da
>   id 00:00:00:00:00:00 priority 32768 hellotime 2 fwddelay 15
>   maxage 20 holdcnt 6 proto rstp maxaddr 100 timeout 1200
>   root id 00:00:00:00:00:00 priority 32768 ifcost 0 port 0
>   member: tap0 flags=143
>   ifmaxaddr 0 port 11 priority 128 path cost 200
>   member: em1 flags=143
>   ifmaxaddr 0 port 2 priority 128 path cost 2
> tap0: flags=8943 metric 
> 0 mtu 1500
>   ether 00:bd:48:03:00:00
>   Opened by PID 24616
> carp1: flags=49 metric 0 mtu 1500
>   inet 10.0.80.74 netmask 0xff00
>   carp: MASTER vhid 2 advbase 1 advskew 0
> 
> 
> The problem I have is that when I ping the VIP from a VPN client (on 
> tap0), the server receives arp requests for the VIP on tap0, but it does 
> not respond to them:
> 
> # tcpdump -i tap0 -ln
> 11:29:13.637048 arp who-has 10.0.80.74 tell 10.0.80.6
> 
> Is there any way to get the server to respond to arp requests on tap0 
> for the VIP?
> 

Maybe you've to do ARP Proxy on one side ? Try to add an ARP entry in
the ARP table with arp (arp -s 1.2.3.4 MAC foo) ..

> This is all on FreeBSD 7.1 with OpenVPN 2.0.6 (both client and server).
> 
> Regards,
> Sebastiaan
> 
-- 
Julien Cigar
Belgian Biodiversity Platform
http://www.biodiversity.be
Université Libre de Bruxelles (ULB)
Campus de la Plaine CP 257
Bâtiment NO, Bureau 4 N4 115C (Niveau 4)
Boulevard du Triomphe, entrée ULB 2
B-1050 Bruxelles
Mail: jci...@ulb.ac.be
@biobel: http://biobel.biodiversity.be/person/show/471
Tel : 02 650 57 52

___
freebsd-questions@freebsd.org mailing list
http://lists.freebsd.org/mailman/listinfo/freebsd-questions
To unsubscribe, send any mail to "freebsd-questions-unsubscr...@freebsd.org"


Re: CARP & bridge

2009-04-29 Thread Sebastiaan van Erk

Hi,

Julien Cigar wrote:

On Wed, 2009-04-29 at 11:37 +0200, Sebastiaan van Erk wrote:

Hi,

I have a bridged OpenVPN setup where the OpenVPN tap0 driver is bridged 
(via bridge0) to the physical em1 interface, which has a VIP via a carp1 
interface:


em1: flags=8943 metric 0 
mtu 1500

options=98
ether 00:0c:29:61:2a:55
inet 10.0.80.77 netmask 0xff00 broadcast 10.0.80.255
media: Ethernet autoselect (1000baseTX )
status: active
bridge0: flags=8843 metric 0 mtu 
1500

ether 9a:6a:9f:b2:65:da
id 00:00:00:00:00:00 priority 32768 hellotime 2 fwddelay 15
maxage 20 holdcnt 6 proto rstp maxaddr 100 timeout 1200
root id 00:00:00:00:00:00 priority 32768 ifcost 0 port 0
member: tap0 flags=143
ifmaxaddr 0 port 11 priority 128 path cost 200
member: em1 flags=143
ifmaxaddr 0 port 2 priority 128 path cost 2
tap0: flags=8943 metric 
0 mtu 1500

ether 00:bd:48:03:00:00
Opened by PID 24616
carp1: flags=49 metric 0 mtu 1500
inet 10.0.80.74 netmask 0xff00
carp: MASTER vhid 2 advbase 1 advskew 0


The problem I have is that when I ping the VIP from a VPN client (on 
tap0), the server receives arp requests for the VIP on tap0, but it does 
not respond to them:


# tcpdump -i tap0 -ln
11:29:13.637048 arp who-has 10.0.80.74 tell 10.0.80.6

Is there any way to get the server to respond to arp requests on tap0 
for the VIP?




Maybe you've to do ARP Proxy on one side ? Try to add an ARP entry in
the ARP table with arp (arp -s 1.2.3.4 MAC foo) ..


Thanks for the suggestion.

Ok, static arp works: that is, if I take the carp1 mac address and add 
it to the arp table using:


 arp -s 10.0.80.74 00:00:5e:00:01:02 pub

The ping starts to work. I'm still a bit confused why I have to do this 
though, because I can ping the non-shared IP 10.0.80.77 from the VPN 
client (via tap0) without any static arp, and I can ping the shared VIP 
(10.0.80.74) from clients on the physical network (em1) as well without 
any static arp. It's only when the ping it has to cross the bridge that 
it's an issue.


Regards,
Sebastiaan



smime.p7s
Description: S/MIME Cryptographic Signature


CARP & bridge

2009-04-29 Thread Sebastiaan van Erk

Hi,

I have a bridged OpenVPN setup where the OpenVPN tap0 driver is bridged 
(via bridge0) to the physical em1 interface, which has a VIP via a carp1 
interface:


em1: flags=8943 metric 0 
mtu 1500

options=98
ether 00:0c:29:61:2a:55
inet 10.0.80.77 netmask 0xff00 broadcast 10.0.80.255
media: Ethernet autoselect (1000baseTX )
status: active
bridge0: flags=8843 metric 0 mtu 
1500

ether 9a:6a:9f:b2:65:da
id 00:00:00:00:00:00 priority 32768 hellotime 2 fwddelay 15
maxage 20 holdcnt 6 proto rstp maxaddr 100 timeout 1200
root id 00:00:00:00:00:00 priority 32768 ifcost 0 port 0
member: tap0 flags=143
ifmaxaddr 0 port 11 priority 128 path cost 200
member: em1 flags=143
ifmaxaddr 0 port 2 priority 128 path cost 2
tap0: flags=8943 metric 
0 mtu 1500

ether 00:bd:48:03:00:00
Opened by PID 24616
carp1: flags=49 metric 0 mtu 1500
inet 10.0.80.74 netmask 0xff00
carp: MASTER vhid 2 advbase 1 advskew 0


The problem I have is that when I ping the VIP from a VPN client (on 
tap0), the server receives arp requests for the VIP on tap0, but it does 
not respond to them:


# tcpdump -i tap0 -ln
11:29:13.637048 arp who-has 10.0.80.74 tell 10.0.80.6

Is there any way to get the server to respond to arp requests on tap0 
for the VIP?


This is all on FreeBSD 7.1 with OpenVPN 2.0.6 (both client and server).

Regards,
Sebastiaan



smime.p7s
Description: S/MIME Cryptographic Signature