Re: [pfSense] fast CF cards?

2012-11-07 Thread Chris Bagnall
On 6 Nov 2012, at 19:24, David Burgess apt@gmail.com wrote:
 With that in mind, can anybody recommend a CF card with good write speed and 
 good reliability?

We've used a mix of Sandisk, Transcend and Kingston cards over the years. Of 
those:
  - nearly all the Kingston cards have failed sooner or later - some after just 
a few months, some after several years
  - we've only had 1 Transcend card fail (out of a few dozen)
  - none of the Sandisk cards have failed

It's difficult to give an accurate view regarding speed, as many of the cards 
in our sample have been bought over several years, and CF cards have tended to 
get faster in recent years (so comparing a 5 year old Sandisk with a brand new 
Transcend isn't really fair).

I will say that Sandisk cards are much 'truer' to their rated speed than the 
other two. I did some tests a few months ago (admittedly for photography rather 
than pfSense) and I found that the Transcend and Duracell 600x cards were 
less than half their rated speeds.

On the other hand, Transcend cards are usually available for less than 10 GBP, 
which if you're ordering lots of them, is a consideration.

Kind regards,

Chris
-- 
This email is made from 100% recycled electrons

___
List mailing list
List@lists.pfsense.org
http://lists.pfsense.org/mailman/listinfo/list


Re: [pfSense] fast CF cards?

2012-11-07 Thread Jim Thompson


On Nov 7, 2012, at 1:59 AM, Chris Bagnall pfse...@lists.minotaur.cc wrote:

 On the other hand, Transcend cards are usually available for less than 10 
 GBP, which if you're ordering lots of them, is a consideration.

We order a lot of CF (1,000 at a time), we don't buy Transcend or on price 
alone. 

We've also never had a Kingston CF fail that I know of. 

Jim
___
List mailing list
List@lists.pfsense.org
http://lists.pfsense.org/mailman/listinfo/list


Re: [pfSense] fast CF cards?

2012-11-07 Thread David Burgess
On Wed, Nov 7, 2012 at 9:46 AM, Jim Thompson j...@netgate.com wrote:


 We've also never had a Kingston CF fail that I know of.


Thanks, everybody, for the feedback. I settled on a Sandisk 200x 8GB. There
were some Kingston's available with much faster ratings, but after reading
some reviews of them, it seems the Kingston are often much slower than they
claim. I've also read that Kingston has a reputation for buying whatever
flash happens to be available, resulting in poor consistency. I'd just
rather stick to something with consistently good reviews, and Sandisk
appears to fit that bill.

For the record, my current CF is a Lexar Pro 2GB. Reliability has been
impeccable in the current build, but not so much in a net5501. I have a
Lexar Pro 4GB in the net5501 currently, and it seems to be fine after a few
years. I can't comment on the speed of either of these, as I've never done
any objective testing; all I can say is that anything less than an instant
firmware update is too slow. ;)

db
___
List mailing list
List@lists.pfsense.org
http://lists.pfsense.org/mailman/listinfo/list


[pfSense] Question about accessing two pfSense boxes in Fail-over mode

2012-11-07 Thread j...@millican.us

Hello,
I know this is a bit short on details but...  I have 4 pfSense boxes in 
two fail-over sets, one set is my edge firewall and the other is inside 
of the first between LAN and a DB zone.  I have remote access through 
OpenVPN that puts me in the LAN where I can get to interface IP's of 
either set of pfSense boxes.  On the boxes between the LAN and the DB 
zone I can connect to either box using what would normally be either of 
the WAN interface IP's.  The problem is that on the edge boxes I can 
only get to the primary, the slave is inaccessible.  The only difference 
I can see is which zone the interface I am trying to access is in, WAN 
vs LAN. The access rules are the same on both sets.  Is there some 
reason that would prevent me from accessing both pfSense boxes while 
they are in fail over mode from the LAN side as I have described?  I am 
wondering if it is because while a master is up the slave just doesn't 
respond to traffic on the LAN side but does on the WAN side?  Has anyone 
else run into a situation like this?  If so could you share your 
solution if you found one?  I know I could create a vpn to both of the 
WAN IP's of the Edge FW's but I would like to limit access to just the 
one Carp IP into the LAN Zone.

Thanks,
JohnM

___
List mailing list
List@lists.pfsense.org
http://lists.pfsense.org/mailman/listinfo/list


Re: [pfSense] Question about accessing two pfSense boxes in Fail-over mode

2012-11-07 Thread Jim Pingle
On 11/7/2012 12:33 PM, j...@millican.us wrote:
 The problem is that on the edge boxes I can
 only get to the primary, the slave is inaccessible.  

In this case, it's likely that your slave box has a route or IPsec phase
2 defined that covers your client subnet, so the slave thinks it knows
the way back to the client directly and the traffic dies because it
really doesn't.

Easiest way around it is manual outbound NAT on the LAN interface to
make traffic going to the secondary appear to originate from the
primary's LAN IP (on LAN, source = VPN subnet, destination = secondary's
IP, translated to Interface address -- NOT the CARP VIP)

If it's OpenVPN, on 2.0.2 and 2.1, binding the VPN to the CARP VIP will
make the server process stop on the backup unit, so the route wouldn't
be maintained in this case, so it should work fine there, so long as the
VPN is bound to a CARP VIP. When the VIP transitions to master it starts
the VPN processes.

Jim

___
List mailing list
List@lists.pfsense.org
http://lists.pfsense.org/mailman/listinfo/list


Re: [pfSense] Question about accessing two pfSense boxes in Fail-over mode

2012-11-07 Thread j...@millican.us

On 11/7/2012 1:23 PM, Jim Pingle wrote:

On 11/7/2012 12:33 PM, j...@millican.us wrote:

The problem is that on the edge boxes I can
only get to the primary, the slave is inaccessible.

In this case, it's likely that your slave box has a route or IPsec phase
2 defined that covers your client subnet, so the slave thinks it knows
the way back to the client directly and the traffic dies because it
really doesn't.

Easiest way around it is manual outbound NAT on the LAN interface to
make traffic going to the secondary appear to originate from the
primary's LAN IP (on LAN, source = VPN subnet, destination = secondary's
IP, translated to Interface address -- NOT the CARP VIP)

If it's OpenVPN, on 2.0.2 and 2.1, binding the VPN to the CARP VIP will
make the server process stop on the backup unit, so the route wouldn't
be maintained in this case, so it should work fine there, so long as the
VPN is bound to a CARP VIP. When the VIP transitions to master it starts
the VPN processes.

Jim


Jim,
Thanks for the response.   I am using OpenVPN with pfSense 
2.0.1-RELEASE** (amd64). I have both boxes using public IPs on the WAN.  
I VPN into the private address space of the LAN (call it 192.168.0.0/24) 
via a public CARP IP which gets my Laptop an address of say 192.168.10 5 
and an address of 192.168.0.1 at the OpenVPN end of the tunnel. OpenVPN 
then routing to the 192.168.0.0/24 network at both ends(added all this 
just for clarity of my statements).  When I open a browser to 
192.168.0.1 (LAN ip of the master FW) all is good.  When I try to browse 
to 192.168.0.2(LAN IP of the slave FW) I get nothing.  I guess I am 
being a bit thick here but I do not understand your last paragraph.   
When you say binding the VPN to the CARP VIP will make the server 
process stop on the backup unit are you referring to the OpenVPN Server 
process?  If so, that should not effect my ability to connect to the 
192.168.0.2 IP, correct? Unfortunately though I can not connect to that 
IP.  I am probably missing something simple but I am at a loss as to 
what. The two DB firewalls have their WAN interface in the 
192.168.0.0/24 net (I.E 192.168.0.253 and 192.168.0.254) with a CARP IP 
(192.168.0.252) that is used to connect from the LAN servers to the DB 
servers.  I can get to the web configuration page on both 192.168.0.254 
and 192.168.0.253 all the time.  I am only having trouble connecting to 
the 192.168.0.2 address for the web configuration page of the 
backup/slave edge FW unit.

Thanks again for that last response,
JohnM

___
List mailing list
List@lists.pfsense.org
http://lists.pfsense.org/mailman/listinfo/list


[pfSense] Bug in pfSense v2.1

2012-11-07 Thread Oliver Schad
Hi all,

I've found a bug in the latest development version v2.1

If you use a carp device a NAT rule is generated which source nats any
outgoing packet to the carp IP.

You can do that if the device is in master mode but you shouldn't do
this if the device is in the backup mode.

The rule is active in both cases. This results in a wrong return path
for all outgoing packets on the backup device (which returns all to
the master device).

In my opinion this shouldn't happen and is a bug.

Regards
Oli


signature.asc
Description: PGP signature
___
List mailing list
List@lists.pfsense.org
http://lists.pfsense.org/mailman/listinfo/list


[pfSense] IPv6-Bug in pfSense v2.1

2012-11-07 Thread Oliver Schad
Hi all,

if I use CARP-Devices with IPv6 I see the following notice:

[ There were error(s) loading the rules: no IP address found
for 2001:abcd:abcd:201::1...0/64/tmp/rules.debug:153: could not
parse host specificationno IP address found for
2001:abcd:abcd:200::2...0/64/tmp/rules.debug:154: could not
parse host specificationpfctl: Syntax error in config file: pf
rules not loaded - The line in question reads [8]: loopback =
{ lo0 }]

I think this is a bug. Do you agree?

Regards
Oli


signature.asc
Description: PGP signature
___
List mailing list
List@lists.pfsense.org
http://lists.pfsense.org/mailman/listinfo/list