ProxyARP Subnetting HOWTO

Bob Edwards

             [email protected]

  v2.0, 27 August 2000

  This HOWTO discusses using Proxy Address Resolution Protocol (ARP)
  with subnetting in order to make a small network of machines visible
  on another Internet Protocol (IP) subnet (I call it sub-subnetting).
  This makes all the machines on the local network (network 0 from now
  on) appear as if they are connected to the main network (network 1).

  This is only relevent if all machines are connected by Ethernet or
  ether devices (ie. it won't work for SLIP/PPP/CSLIP etc.)
    _________________________________________________________________

  Table of Contents
  1. [1]Acknowledgements
  2. [2]Why use Proxy ARP with subnetting?
  3. [3]How Proxy ARP with subnetting works
  4. [4]Setting up Proxy ARP with subnetting
  5. [5]Other alternatives to Proxy ARP with subnetting
  6. [6]Other Applications of Proxy ARP with subnetting
  7. [7]Copying conditions

1. Acknowledgements

  This document, and my Proxy ARP implementation could not have been
  made possible without the help of:

    * Andrew Tridgell, who implemented the subnetting options for arp in
      Linux, and who personally assisted me in getting it working
    * the Proxy-ARP mini-HOWTO, by Al Longyear
    * the Multiple-Ethernet mini-HOWTO, by Don Becker
    * the arp(8) source code and man page by Fred N. van Kempen and
      Bernd Eckenfels
    _________________________________________________________________

2. Why use Proxy ARP with subnetting?

  The applications for using Proxy ARP with subnetting are fairly
  specific.

  In my case, I had a wireless Ethernet card that plugs into an 8-bit
  ISA slot. I wanted to use this card to provide connectivity for a
  number of machines at once. Being an ISA card, I could use it on a
  Linux machine, after I had written an appropriate device driver for it
  - this is the subject of another document. From here, it was only
  necessary to add a second Ethernet interface to the Linux machine and
  then use some mechanism to join the two networks together.

  For the purposes of discussion, let network 0 be the local Ethernet
  connected to the Linux box via an NE-2000 clone Ethernet interface on
  eth0. Network 1 is the main network connected via the wireless
  Ethernet card on eth1. Machine A is the Linux box with both
  interfaces. Machine B is any TCP/IP machine on network 0 and machine C
  is likewise on network 1.

  Normally, to provide the connectivity, I would have done one of the
  following:

    * Used the IP-Bridge software (see the Bridge mini-HOWTO) to bridge
      the traffic between the two network interfaces. Unfortunately, the
      wireless Ethernet interface cannot be put into "Promiscuous" mode
      (ie. it can't see all packets on network 1). This is mainly due to
      the lower bandwidth of the wireless Ethernet (2MBit/sec) meaning
      that we don't want to carry any traffic not specifically destined
      to another wireless Ethernet machine - in our case machine A - or
      broadcasts. Also, bridging is rather CPU intensive!
    * Alternatively, use subnets and an IP-router to pass packets
      between the two networks (see the IP-Subnetworking mini-HOWTO).
      This is a protocol specific solution, where the Linux kernel can
      handle the Internet Protocol (IP) packets, but other protocols
      (such as AppleTalk) need extra software to route. This also
      requires the allocation of a new IP subnet (network) number, which
      is not always an option.

  In my case, getting a new subnet (network) number was not an option,
  so I wanted a solution that allowed all the machines on network 0 to
  appear as if they were on network 1. This is where Proxy ARP comes in.
  Other solutions are used to connect other (non-IP) protocols, such as
  netatalk to provide AppleTalk routing.
    _________________________________________________________________

3. How Proxy ARP with subnetting works

  The Proxy ARP is actually only used to get packets from network 1 to
  network 0. To get packets back the other way, the normal IP routing
  functionality is employed.

  In my case, network 1 has an 8-bit subnet mask (255.255.255.0). I have
  chosen the subnet mask for network 0 to be 4-bit (255.255.255.240),
  allowing 14 IP nodes on network 0 (2 ^ 4 = 16, less two for the all
  zeros and all ones cases). Note that any size of subnet mask up to,
  but not including, the size of the mask of the other network is
  allowable here (eg. 2, 3, 4, 5, 6 or 7 bits in this case - for one
  bit, just use normal Proxy ARP!)

  All the IP numbers for network 0 (16 in total) appear in network 1 as
  a subset. Note that it is very important, in this case, not to allow
  any machine connected directly to network 1 to have an IP number in
  this range! In my case, I have "reserved" the IP numbers of network 1
  ending in 64 .. 79 for network 0. In this case, the IP numbers ending
  in 64 and 79 can't actually be used by nodes - 79 is the broadcast
  address for network 0.

  Machine A is allocated two IP numbers, one within the network 0 range
  for it's real Ethernet interface (eth0) and the other within the
  network 1 range, but outside of the network 0 range, for the wireless
  Ethernet interface (eth1).

  Say machine C (on network 1) wants to send a packet to machine B (on
  network 0). Because the IP number of machine B makes it look to
  machine C as though it is on the same physical network, machine C will
  use the Address Resolution Protocol (ARP) to send a broadcast message
  on network 1 requesting the machine with the IP number of machine B to
  respond with it's hardware (Ethernet or MAC layer) address. Machine B
  won't see this request, as it isn't actually on network 1, but machine
  A, on both networks, will see it.

  The first bit of magic now happens as the Linux kernel arp code on
  machine A, with a properly configured Proxy ARP with subnetting entry,
  determines that the ARP request has come in on the network 1 interface
  (eth1) and that the IP number being ARP'd for is in the subnet range
  for network 0. Machine A then sends it's own hardware (Ethernet)
  address back to machine C as an ARP response packet.

  Machine C then updates it's ARP cache with an entry for machine B, but
  with the hardware (Ethernet) address of machine A (in this case, the
  wireless Ethernet interface). Machine C can now send the packet for
  machine B to this hardware (Ethernet) address, and machine A receives
  it.

  Machine A notices that the destination IP number in the packet is that
  of machine B, not itself. Machine A's Linux kernel IP routing code
  attempts to forward the packet to machine B by looking at it's routing
  tables to determine which interface contains the network number for
  machine B. However, the IP number for machine B is valid for both the
  network 0 interface (eth0), and for the network 1 interface (eth1).

  At this point, something else clever happens. Because the subnet mask
  for the network 0 interface has more 1 bits (it is more specific) than
  the subnet mask for the network 1 interface, the Linux kernel routing
  code will match the IP number for machine B to the network 0
  interface, and not keep looking for the potential match with the
  network 1 interface (the one the packet came in on).

  Now machine A needs to find out the "real" hardware (Ethernet) address
  for machine B (assuming that it doesn't already have it in the ARP
  cache). Machine A uses an ARP request, but this time the Linux kernel
  arp code notes that the request isn't coming from the network 1
  interface (eth1), and so doesn't respond with the Proxy address of
  eth1. Instead, it sends the ARP request on the network 0 interface
  (eth0), where machine B will see it and respond with it's own (real)
  hardware (Ethernet) address. Now machine A can send the packet (from
  machine C) onto machine B.

  Machine B gets the packet from machine C (via machine A) and then
  wants to send back a response. This time, machine B notices that
  machine C in on a different subnet (machine B's subnet mask of
  255.255.255.240 excludes all machines not in the network 0 IP address
  range). Machine B is setup with a "default" route to machine A's
  network 0 IP number and sends the packet to machine A. This time,
  machine A's Linux kernel routing code determines the destination IP
  number (of machine C) as being on network 1 and sends the packet onto
  machine C via Ethernet interface eth1.

  Similar (less complicated) things occur for packets originating from
  and destined to machine A from other machines on either of the two
  networks.

  Similarly, it should be obvious that if another machine (D) on network
  0 ARP's for machine B, machine A will receive the ARP request on it's
  network 0 interface (eth0) and won't respond to the request as it is
  set up to only Proxy on it's network 1 interface (eth1).

  Note also that all of machines B and C (and D) are not required to do
  anything unusual, IP-wise. In my case, there is a mixture of Suns,
  Macs and PC/Windoze 95 machines on network 0 all connecting through
  Linux machine A to the rest of the world.

  Finally, note that once the hardware (Ethernet) addresses are
  discovered by each of machines A, B, C (and D), they are placed in the
  ARP cache and subsequent packet transfers occur without the ARP
  overhead. The ARP caches normally expire entries after 5 minutes of
  non-activity.
    _________________________________________________________________

4. Setting up Proxy ARP with subnetting

  I set up Proxy ARP with subnetting on a Linux kernel version 2.0.30
  machine, but I am told that the code works right back to some kernel
  version in the 1.2.x era.

  The first thing to note is that the ARP code is in two parts: the part
  inside the kernel that sends and receives ARP requests and responses
  and updates the ARP cache etc.; and other part is the arp(8) command
  that allows the super user to modify the ARP cache manually and anyone
  to examine it.

  The first problem I had was that the arp(8) command that came with my
  Slackware 3.1 distribution was ancient (1994 era!!!) and didn't
  communicate with the kernel arp code correctly at all (mainly
  evidenced by the strange output that it gave for "arp -a").

  The arp(8) command in "net-tools-1.33a" available from a variety of
  places, including (from the README file that came with it)
  [8]ftp.linux.org.uk:/pub/linux/Networking/base/ works properly and
  includes new man pages that explain stuff a lot better than the older
  arp(8) man page.

  Armed with a decent arp(8) command, all the changes I made were in the
  /etc/rc.d/rc.inet1 script (for Slackware - probably different for
  other flavours). First of all, we need to change the broadcast
  address, network number and netmask of eth0:

NETMASK=255.255.255.240 # for a 4-bit host part
NETWORK=x.y.z.64        # our new network number (replace x.y.z with your net)
BROADCAST=x.y.z.79      # in my case

  Then a line needs to be added to configure the second Ethernet port
  (after any module loading that might be required to load the driver
  code):

/sbin/ifconfig eth1 (name on net 1) broadcast (x.y.z.255) netmask 255.255.255.0

  Then we add a route for the new interface:

/sbin/route add -net (x.y.z.0) netmask 255.255.255.0

  And you will probably need to change the default gateway to the one
  for network 1.

  At this point, it is appropriate to add the Proxy ARP entry:

/sbin/arp -i eth1 -Ds ${NETWORK} eth1 netmask ${NETMASK} pub

  This tells ARP to add a static entry (the s) to the cache for network
  ${NETWORK}. The -D tells ARP to use the same hardware address as
  interface eth1 (the second eth1), thus saving us from having to look
  up the hardware address for eth1 and hardcoding it in. The netmask
  option tells ARP that we want to use subnetting (ie. Proxy for all (IP
  number) & ${NETMASK} == ${NETWORK} & ${NETMASK}). The pub option tells
  ARP to publish this ARP entry, ie. it is a Proxy entry, so respond on
  behalf of these IP numbers. The -i eth1 option tells ARP to only
  respond to requests that come in on interface eth1.

  Hopefully, at this point, when the machine is rebooted, all the
  machines on network 0 will appear to be on network 1. You can check
  that the Proxy ARP with subnetting entry has been correctly installed
  on machine A. On my machine (names changed to protect the innocent) it
  is:

bash$ /sbin/arp -an
Address                 HWtype  HWaddress           Flags Mask            Iface
x.y.z.1                 ether   00:00:0C:13:6F:17   C     *               eth1
x.y.z.65                ether   00:40:05:49:77:01   C     *               eth0
x.y.z.67                ether   08:00:20:0B:79:47   C     *               eth0
x.y.z.5                 ether   00:00:3B:80:18:E5   C     *               eth1
x.y.z.64                ether   00:40:96:20:CD:D2   CMP   255.255.255.240 eth1

  Alternatively, you can examine the /proc/net/arp file with eg. cat(1).

  The last line is the proxy entry for the subnet. The CMP flags
  indicate that it is a static (Manually entered) entry and that it is
  to be Published. The entry is only going to reply to ARP requests on
  eth1 where the requested IP number, once masked, matches the network
  number, also masked. Note that arp(8) has automatically determined the
  hardware address of eth1 and inserted this for the address to use (the
  -Ds option).

  Likewise, it is probably prudent to check that the routing table has
  been set up correctly. Here is mine (again, the names are changed to
  protect the innocent):

#/bin/netstat -rn
Kernel routing table
Destination     Gateway         Genmask         Flags Metric Ref Use    Iface
x.y.z.64        0.0.0.0         255.255.255.240 U     0      0       71 eth0
x.y.z.0         0.0.0.0         255.255.255.0   U     0      0      389 eth1
127.0.0.0       0.0.0.0         255.0.0.0       U     0      0        7 lo
0.0.0.0         x.y.z.1         0.0.0.0         UG    1      0      573 eth1

  Alternatively, you can examine the /proc/net/route file with eg.
  cat(1).

  Note that the first entry is a proper subset of the second, but the
  routing table has ranked them in netmask order, so the eth0 entry will
  be checked before the eth1 entry.
    _________________________________________________________________

5. Other alternatives to Proxy ARP with subnetting

  There are several other alternatives to using Proxy ARP with
  subnetting in this situation, apart from the ones mentioned about
  (bridging and straight routing):

    * IP-Masquerading (see the IP-Masquerade mini-HOWTO), in which
      network 0 is "hidden" behind machine A from the rest of the
      Internet. As machines on network 0 attempt to connect outside
      through machine A, it re-addresses the source address and port
      number of the packets and makes them look like they are coming
      from itself, rather than from the machine on the hidden network 0.
      This is an elegant solution, although it prevents any machine on
      network 1 from initiating a connection to any machine on network
      0, as the machines on network 0 effectively don't exist outside of
      network 0. This effectively increases security of the machines on
      network 0, but is also means that servers on network 1 cannot
      check the identity of clients on network 0 using IP numbers (eg.
      NFS servers use IP hostnames for access to mountable file
      systems).
    * Another option is IP in IP tunneling, which isn't supported on all
      platforms (such as Macs and Windoze machines) so I opted not to go
      this way.
    * Use Proxy ARP without subnetting. This is certainly possible, it
      just means that a separate entry needs to be created for each
      machine on network 0, instead of a single entry for all machines
      (current and future) on network 0.
    * Possibly IP Aliasing might also be useful here, but I haven't
      looked into this at all.
    _________________________________________________________________

6. Other Applications of Proxy ARP with subnetting

  There is only one other application that I know about that uses Proxy
  ARP with subnetting, also here at the Australian National University.
  It is the one that Andrew Tridgell originally wrote the subnetting
  extensions to Proxy ARP for. However, Andrew reliably informs me that
  there are, in fact, several other sites around the world using it as
  well (I don't have any details).

  The other A.N.U. application involves a teaching lab set up to teach
  students how to configure machines to use TCP/IP, including setting up
  the gateway. The network used is a Class C network, and Andrew needed
  to "subnet" it for security, traffic control and the educational
  reason mentioned above. He did this using Proxy ARP, and then decided
  that a single entry in the ARP cache for the whole subnet would be
  faster and cleaner than one for each host on the subnet. Voila...Proxy
  ARP with subnetting!
    _________________________________________________________________

7. Copying conditions

  Copyright 1997 by Bob Edwards <[9][email protected]>

  Voice: (+61) 2 6249 4090

  Unless otherwise stated, Linux HOWTO documents are copyrighted by
  their respective authors. Linux HOWTO documents may be reproduced and
  distributed in whole or in part, in any medium physical or electronic,
  as long as this copyright notice is retained on all copies. Commercial
  redistribution is allowed and encouraged; however, the author would
  like to be notified of any such distributions. All translations,
  derivative works, or aggregate works incorporating any Linux HOWTO
  documents must be covered under this copyright notice. That is, you
  may not produce a derivative work from a HOWTO and impose additional
  restrictions on its distribution. Exceptions to these rules may be
  granted under certain conditions; please contact the Linux HOWTO
  coordinator at the address given below. In short, we wish to promote
  dissemination of this information through as many channels as
  possible. However, we do wish to retain copyright on the HOWTO
  documents, and would like to be notified of any plans to redistribute
  the HOWTOs. If you have questions, please contact the Linux HOWTO
  coordinator, at <[10][email protected]> via email.

References

  1. Proxy-ARP-Subnet.html#INTRO
  2. Proxy-ARP-Subnet.html#WHY
  3. Proxy-ARP-Subnet.html#HOW
  4. Proxy-ARP-Subnet.html#SETUP
  5. Proxy-ARP-Subnet.html#ALTERNATIVES
  6. Proxy-ARP-Subnet.html#APPLICATIONS
  7. Proxy-ARP-Subnet.html#COPYING
  8. ftp://ftp.linux.org.uk/pub/linux/Networking/base/
  9. mailto:[email protected]
 10. mailto:[email protected]

You are viewing proxied material from ftp.icm.edu.pl. The copyright of proxied material belongs to its original authors. Any comments or complaints in relation to proxied material should be directed to the original authors of the content concerned. Please see the disclaimer for more details.