Embedded Linux VPN Router Near Release
An anonymous reader writes "A new open source project aims to build a VPN router that supports all major routing protocols on a standardized hardware platform running embedded Linux. The "Linux Router Project - LR101" started in mid-2003 and plans a first release in January 2004. It is based on a dual-NIC VIA EPIA mainboard and a Travla case, along with Red Hat 8, zebra, FreeS/WAN, IP-tables, an other open source software, all compiled from source."
It would be nice if they have High Availability on their feature list. Some nice solid appliances like this would be interesting.
Is this a stripped down Redhat distro, with a configuration tool that they wrote? Isn't a whole distribution a little bit too much for such a project? Wouldn't a linuxfromscratch installation - with only the bare minimums - be a better idea? Just a thought.
Where's PPTP? for a VPN router, it's kind of desirable ...
"A door is what a dog is perpetually on the wrong side of" - Ogden Nash
Or, just buy a Linux-based Linksys WRV54G for well under $200 with most, if not all the features of this project. No, I don't mean the WRT54g, I mean the WRV54G. Excellent piece of gear, VPN, firewalling, dmz, wireless (wep/wpa), snmp, yadda yadda.
If that's true, then it's illegal for a US citizen to contribute to the 2.6.0 kernel too, since that has crypto in it.
all compiled from source.
As opposed to say, a Linksys Router, which we all know is compiled from Cheerios. =)
Do you Gentoo!?
Soekris
...the Linux Router Project, a floppy-based 386-compatible micro-distro which served as the basis for (among other things) Coyote Linux.
Custom firmware for the wrt54g does/will do pretty much the same thing. Progress is very quick. See the forum here:
sveasoft
The initial PPTP was a total botch, with seven major security flaws. Some of them have since been fixed, but it gives you some idea of the professionalism and quality that didn't go into the basic design. If you want to use a VPN for security, use IPSEC - and this project has FreeS/WAN IPSEC in it. If you really really want to use a VPN to transport lame non-IP legacy Microsoft LAN protocols, go pay Microsoft some money for one of their server projects, and charge the silly customer who's hiring you as a consultant because they don't want to upgrade to the 1990s for it. If you want to use a VPN to carry private IP addresses, but don't actually care about security, use IPSEC anyway, or use GRE tunnels.
Bill Stewart
New Fast-Compression-only CPR http://preview.tinyurl.com/dy575ks
Indeed, export of cryptographic technology from USA is hampered with strong restrictions. So many Open Source projects are quite careful to avoid breaking laws by having (much) development done outside USA, and also letting release builds be done outside US as well.
For instance, OpenBSD has offered strong encryption for several years. The OpenBSD project is located in Canada, and a lot of development/release builds are done outside US. As Integrated Crypto shows :
RealTek is RealCrap. You get what you pay for.
/usr/src/sys/pci/if_rl.c on my FreeBSD system:
From
* The RealTek 8139 PCI NIC redefines the meaning of 'low end.' This is
* probably the worst PCI ethernet controller ever made, with the possible
* exception of the FEAST chip made by SMC. The 8139 supports bus-master
* DMA, but it has a terrible interface that nullifies any performance
* gains that bus-master DMA usually offers.
*
* It's impossible given this rotten design to really achieve decent
* performance at 100Mbps, unless you happen to have a 400Mhz PII or
* some equally overmuscled CPU to drive it.
This is my favorite comment:
* Here's a totally undocumented fact for you. When the
* RealTek chip is in the process of copying a packet into
* RAM for you, the length will be 0xfff0. If you spot a
* packet header with this value, you need to stop. The
* datasheet makes absolutely no mention of this and
* RealTek should be shot for this.
More funny stuff:
* The RealTek is brain damaged and wants longword-aligned
* TX buffers, plus we can only have one fragment buffer
* per packet. We have to copy pretty much all the time.