Linux IPX-HOWTO
 Kevin Thorpe, [email protected]
 v2.3, 06 May 1998

 This document aims to describe how to obtain, install and configure
 various tools available for the the Linux operating system that use
 the Linux kernel IPX protocol support.
 ______________________________________________________________________

 Table of Contents



 1. Introduction.

    1.1 Changes from the previous release.
    1.2 Introduction.

 2. Disclaimer.

 3. Related Documentation.

    3.1 New versions of this document.
    3.2 Feedback.
    3.3 Mailing list support.

 4. Some of the terms used in this document.

 5. A brief discussion of IPX network topology

 6. The IPX related files in the

 7. Greg Pages IPX tools.

    7.1 The IPX tools in more detail.

 8. Configuring your Linux machine as an IPX router.

    8.1 Do I need to configure an internal network ?

 9. Configuring your Linux machine as an NCP client.

    9.1 Obtaining
    9.2 Building
    9.3 Building
    9.4 Configuring and using

 10. Configuring your Linux machine as an NCP server.

    10.1 The
       10.1.1 Capability of
       10.1.2 Obtaining
       10.1.3 Building the
    10.2 The
       10.2.1 Capability of
       10.2.2 Obtaining
       10.2.3 Building
       10.2.4 Configuring and using

 11. Configuring your Linux machine as a Novell Print Client.

 12. Configuring your Linux machine as a Novell Print Server.

    12.1 Prerequisites
    12.2 Configuration

 13. An overview of the

    13.1 User commands.
    13.2 Administration tools.

 14. Configuring PPP for IPX support.

    14.1 Configuring an IPX/PPP server.
       14.1.1 First steps.
       14.1.2 Design.
       14.1.3 Configure
       14.1.4 Test the server configuration.
    14.2 Configuring an IPX/PPP client.
       14.2.1 Configuring
       14.2.2 Testing the IPX/PPP client.

 15. IPX tunnel over IP

    15.1 Obtaining
    15.2 Building
    15.3 Configuring
    15.4 Testing and using

 16. Commercial IPX support for Linux.

    16.1 Caldera'a Network Desktop

 17. Some Frequently Asked Questions

 18. Copyright Message.

 19. Miscellaneous and Acknowledgements.



 ______________________________________________________________________

 1.  Introduction.

 This is the Linux IPX-HOWTO. You should read the Linux NET-3-HOWTO in
 conjunction with this document.


 1.1.  Changes from the previous release.



 Change of author:
         Many thanks to Terry Dawson for passing on this document and
         congratulations on becoming a father :-).

 Additions:
         Addition of a brief explanation of IPX. This is in response to
         many baffled queries on the discussion lists.

 Corrections/Updates:
         New version of ncpfs which now supports NDS logins. This is early
         beta test and may be prohibited in your country due to the use of
         patented technology.

         Addition of support for trustee rights in mars_nwe. This is still
         in beta test.



 1.2.  Introduction.

 The Linux Kernel has a completely new network implementation as
 compared to other Unix like operating systems. The ability to take a
 fresh approach to developing the kernel networking software has led to
 the Linux kernel having support for a range of non tcp/ip protocols
 being built. The IPX protocol is one of those that have been included.

 The Linux kernel supports the IPX protocol only. It does not yet
 support protocols such as IPX/RIP, SAP or NCP, these are supported by
 other software such as that documented elsewhere in this document.


 The IPX support was originally developed by Alan Cox
 <[email protected]> and has been significantly enhanced by Greg
 Page <[email protected]>.


 2.  Disclaimer.

 I do not and cannot know everything there is to know about the Linux
 network software. Please accept and be warned that this document
 probably does contain errors. Please read any README files that are
 included with any of the various pieces of software described in this
 document for more detailed and accurate information. I will attempt to
 keep this document as error-free and up-to-date as possible. Versions
 of software are current as at time of writing.

 In no way do I or the authors of the software in this document offer
 protection against your own actions. If you configure this software,
 even as described in this document and it causes problems on your
 network then you alone must carry the responsibility. I include this
 warning because IPX network design and configuration is not always a
 simple matter and sometimes undesirable interaction with other routers
 and fileservers can result if you do not design or configure your
 network carefully. I also include this warning because I was asked to
 by someone unfortunate enough to have discovered this lesson the hard
 way.


 3.  Related Documentation.

 This document presumes you understand how to build a Linux kernel with
 the appropriate networking options selected and that you understand
 how to use the basic network tools such as ifconfig and route.  If you
 do not, then you should read the NET-3-HOWTO <NET-3-HOWTO.html> in
 conjunction with this document as it describes these.

 Other Linux HOWTO documents that might be useful are:

 The Ethernet-HOWTO <Ethernet-HOWTO.html>, which describes the details
 of configuring an Ethernet device for Linux.

 The PPP-HOWTO <PPP-HOWTO.html> as IPX support is available for version
 2.2.0d and later of the Linux PPP implementation.


 3.1.  New versions of this document.

 If your copy of this document is more than two months old then I
 strongly recommend you obtain a newer version. The networking support
 for Linux is changing very rapidly with new enhancements and features,
 so this document also changes fairly frequently. The latest released
 version of this document can always be retrieved by anonymous ftp
 from:

 ftp:/sunsite.unc.edu/pub/Linux/docs/HOWTO/IPX-HOWTO>/ or:
 ftp:/sunsite.unc.edu/pub/Linux/docs/HOWTO/other-formats/IPX-
 HOWTO{-html.tar,ps,dvi}.gz>/ via the World Wide Web from the Linux
 Documentation Project Web Server
 <http://sunsite.unc.edu/LDP/linux.html>, at page: IPX-HOWTO
 <http://sunsite.unc.edu/LDP/HOWTO/IPX-HOWTO.html> or directly from me,
 <[email protected]>. It may also be posted to the newsgroups:
 comp.os.linux.networking, comp.os.linux.answers and news.answers from
 time to time.



 3.2.  Feedback.

 Please send any comments, updates, or suggestions to me,
 <[email protected]>. The sooner I get feedback, the sooner I can
 update and correct this document. If you find any problems with it,
 please mail me directly as I can miss info posted to the newsgroups.


 3.3.  Mailing list support.

 There is a mailing list established for discussion of the various
 Linux IPX software packages described in this document. You can
 subscribe to it by sending a mail message to `[email protected]'
 with `add linware' in the body of the message. To post to the list
 your send your mail to `[email protected]'. I regularly watch this
 list.

 The mailing list is archived at www.kin.vslib.cz
 <http://www.kin.vslib.cz/hypermail/linware/>.


 4.  Some of the terms used in this document.

 You will often see the terms client and server used in this document.
 They are normally fairly specific terms but in this document I have
 generalized their definitions a little so that they mean the
 following:

    client
       The machine or program that initiates an action or a connection
       for the purpose of gaining use of some service or data.

    server
       The machine or program that accepts incoming connections from
       multiple remote machines and provides a service or data to
       those.

 These definitions are not very reliable either, but they provide a
 means of distinguishing the ends of peer to peer systems such as SLIP
 or PPP which truly do not actually have clients and servers.

 Other terms you will see are:

    Bindery
       The bindery is a specialised database storing network
       configuration information on a Novell fileserver. Netware
       clients may query the bindery to obtain information on available
       services, routing and user information.

    Frame Type
       is a term used to describe that actual protocol used to carry
       the IPX (and IP) datagrams across your ethernet style network
       segments. There are four common ones. They are:

       Ethernet_II
          This is a refined version of the original DIX ethernet
          standard. Novell has been allocated a formal protocol id and
          this means that both IPX and IP can coexist happily in an
          Ethernet_II environment quite happily. This is commonly used
          in Novell environments and is a good choice.

       802.3
          This is an I.E.E.E. protocol defining a Carrier Sense
          Multiple Access with Collision Detection (CSMA/CD) mechanism.
          It was based on the original DIX Ethernet standard, with an
          important modification, the type (protocol id) field was
          converted into a length field instead. It is for this reason
          that IPX really shouldn't be run here. IEEE 802.3 was
          designed to carry IEEE 802.2 frames only but there are
          implementations that use it to carry IPX frames directly and
          remarkably it does work. Avoid it unless you are trying to
          interwork with a network already configured to use it.

       802.2
          This is an I.E.E.E. protocol that defines a set of Logical
          Link Control procedures. It provides a simplistic way of
          allowing different protocols to coexist, but is quite limited
          in this respect. Novell uses an unofficial Service Address
          Point (like a protocol id) but since everyone else uses it as
          well, that hasn't yet presented too much of a problem.

       SNAP
          SNAP is the Sub Network Access Protocol. This protocol is
          designed ride on top of 802.3 and 802.2. It expands the
          multiprotocol capability of 802.2 and provides some measure
          of compatability with existing Ethernet and Ethernet_II frame
          types.


    IPX
       Internet Packet eXchange is a protocol used by the Novell
       corporation to provide internetworking support for their
       NetWare(tm) product.  IPX is similar in functionality to the IP
       protocol used by the tcp/ip community.

    IPX network address
       This is a number which uniquely identifies a particular IPX
       network. The usual notation for this address is in hexadecimal.
       An example might look like: 0x23a91002.

    IPX Internal network
       This is a virtual IPX network. It is virtual because it does not
       correspond to a physical network. This is used to provide a
       means of uniquely identifying and addressing a particular IPX
       host. This is generally only useful to IPX hosts that exist on
       more than one physical IPX network such as fileservers. The
       address is coded in the same form as for a physical IPX network.

    RIP
       Routing Information Protocol is a protocol used to automatically
       propagate network routes in an IPX network. It is functionally
       similar to the RIP used within the tcp/ip community.

    NCP
       NetWare Core Protocol is a networked filesystem protocol
       designed by the Novell Corporation for their NetWare(tm)
       product. NCP is functionally similar to the NFS used in the
       tcp/ip community.

    SAP
       Service Advertisement Protocol is a protocol designed by the
       Novell Corporation that is used to advertise network services in
       a NetWare(tm) environment.

    Hardware address
       This is a number that uniquely identifies a host in a physical
       network at the media access layer. Examples of this are Ethernet
       Addresses. An Ethernet address is generally coded as six
       hexadecimal values separated by colon characters eg.
       00:60:8C:C3:3C:0F


    route
       The route is the path that your packets take through the network
       to reach their destination.


 5.  A brief discussion of IPX network topology

 This is a much simplified explanation for people new to IPX. Large
 networks will probably break lots of the rules explained here. In
 complex IPX networks the administrator should always be consulted.

 IPX networking revolves around a scheme of numbered networks unlike IP
 which places more emphasis on the interface addresses. A network is a
 collection of equipment connected to the same LAN segment and using
 the same frame type. Different frame types on the same LAN segment are
 treated as seperate networks.

 Each network must be allocated a number which is unique across the
 entire internetwork. This is usually performed by a NetWare(tm)
 server, but can easily be performed by Linux. IPX clients are given
 this number by the server when starting, they only require to know the
 correct frame type.

 Routing between networks is usually performed by putting two network
 cards in a server. This server then runs the RIP protocol which holds
 a routing table for the internetwork. Periodic broadcasts of this
 routing table are exchanged between servers. Within a short time each
 server 'discovers' the topology of the internetwork.

 If you only wish to use the services of an existing NetWare server,
 you can use ipx_configure (section 7.1) to automatically define the
 IPX interfaces by using broadcast queries to look for a server. If
 this fails, or you wish to provide IPX services, you will need to
 define the interfaces manually using ipx_interface or mars_nwe.


 6.  The IPX related files in the /proc  filesystem.

 There are a number of files related to the Linux IPX support that are
 located within the /proc filesystem. They are:


    /proc/net/ipx_interface
       This file contains information about the IPX interfaces
       configured on your machine. These may have been configured
       manually by command or automatically detected and configured.

    /proc/net/ipx_route
       This file contains a list of the routes that exist in the IPX
       routing table. These routes may have been added manually by
       command or automatically by an IPX routing daemon.

    /proc/net/ipx
       This file is a list of the IPX sockets that are currently open
       for use on the machine.


 7.  Greg Pages IPX tools.

 Greg Page <[email protected] of Caldera Incorporated has written a
 suite of IPX configuration tools and enhanced the Linux IPX kernel
 support.

 The kernel enhancements allow linux to be configured as a fully
 featured IPX bridge or router. The enhanced IPX support has already
 been fed back into the mainstream kernel distribution so you will
 probably already have it.

 The network configuration tools provide you with the capability to
 configure your network devices to support IPX and allow you to
 configure IPX routing and other facilities under Linux. The Linux IPX
 network tools are available from: sunsite.unc.edu
 <ftp://sunsite.unc.edu/pub/Linux/system/filesystems/ncpfs/ipx.tgz>.


 7.1.  The IPX tools in more detail.


    ipx_interface
       This command is used to manually add, delete or check ipx
       capability to an existing network device. Normally the network
       device would be an Ethernet device such at eth0. At least one
       IPX interface must be designated the primary interface and the
       -p flag to this command does this. For example to enable
       Ethernet device eth0 for IPX capability as the primary IPX
       interface using the IEEE 802.2 frame type and IPX network
       address 39ab0222 you would use:


         # ipx_interface add -p eth0 802.2 0x39ab0222



    If the frame type differs from NetWare(tm) servers on this network,
    they will studiously ignore you. If the frame type is correct but
    the network number differs, they will still ignore you but complain
    frequently on the NetWare server console. The latter is guaranteed
    to gain you flames from your NetWare administrator and may disrupt
    existing NetWare clients.

    If you get an error while running this program and you happen to
    not have already configured tcp/ip, then you will find that you
    need to manually start the eth0 interface using the command:


         # ifconfig eth0 up



    ipx_configure
       This command enables or disables the automatic setting of the
       interface configuration and primary interface settings.

       --auto_interface
          allows you to select whether new network devices should be
          automatically configured as IPX devices or not.

       --auto_primary
          allows you to select whether the IPX software should
          automatically select a primary interface or not. Problems
          have been noted using this with Windows 95 clients on the
          network.

       A typical example would be to enable both automatic interface
       configuration and automatic primary interface setting with the
       following command:



    # ipx_configure --auto_interface=on --auto_primary=on



    ipx_internal_net
       This command allows you to configure or deconfigure an internal
       network address. An internal network address is optional, but
       when it is configured it will always be the primary interface.
       To configure an IPX network address of ab000000 on IPX node 1
       you would use:


         # ipx_internal_net add 0xab000000 1



    ipx_route
       The command allows you to manually modify the IPX routing table.
       For example to add a route to IPX network 39ab0222 via a router
       with node number 00608CC33C0F on IPX network 39ab0108:


         # ipx_route add 0x39ab0222 0x39ab0108 0x00608CC33C0F



 8.  Configuring your Linux machine as an IPX router.

 If you have a number of IPX segments that you wish to internetwork you
 need the services of a router. In the Novell environment there are two
 pieces of information which are necessary to be propagated around the
 network.  They are the network routing information propagated using
 Novell RIP, and the service advertisement information propagated using
 Novell SAP. Any router must support both of these protocols to be
 useful in most situations.

 Linux has support for both of these protocols and can be fairly easily
 made to function as a fully Novell compliant router.

 The Linux kernel IPX support actually manages the IPX packet
 forwarding across interfaces, but it does this according to the rules
 coded into the IPX routing table. Linux needs a program to implement
 the Novell RIP and SAP to ensure that the IPX routing table is built
 correctly and updated periodically to reflect changes in the network
 status.

 Volker Lendecke <[email protected]> has developed a routing
 daemon ipxripd that will do this for you. The mars_nwe package
 mentioned later includes an alternative routing daemon.

 You can find ipxripd at:

 sunsite.unc.edu
 <ftp://sunsite.unc.edu/pub/Linux/system/filesystems/ncpfs/ipxripd-0.7.tgz>

 or at Volkers home site at:

 ftp.gwdg.de <ftp://ftp.gwdg.de/pub/linux/misc/ncpfs/ipxripd-0.7.tgz>


 Configuring your Linux machine to act as a router is very
 straightforward.  The steps you must take are:

 1. Build your kernel with IPX, Ethernet and /proc support.

 2. Obtain, compile and install the ipxd daemon program.

 3. Boot the new kernel and ensure that each of the Ethernet cards has
    been properly detected and there are no hardware conflicts.

 4. Enable the IPX protocol on each of the interfaces using the
    ipx_interface command described above.

 5. Start the ipxd daemon program.

 Consider the following simple network:


  IPX Addr: 0x01000000  802.2
 |--------------------------|
               |
               \_________________________
                                         \          Linux Router
  IPX Addr: 0x02000000  802.2             \
 |--------------------------|              \   eth0/-----------\
               |                            \--====|           |
               \_________________________          | IPX route |
                                         \     eth1|  Table    |
  IPX Addr: 0x03000000  etherII           \----====|    ^      |
 |--------------------------|                      |    |      |
               |                               eth2|  IPXd     |
               \______________________________/====|           |
                                                   |  SAPd     |
  IPX Addr: 0x04000000  etherII                eth3|           |
 |--------------------------|                 /====|           |
               |                              |    \___________/
               \______________________________/



 The configuration for the above network would look like:


      # ipx_interface add eth0 802.2 0x0100000000
      # ipx_interface add eth1 802.2 0x0200000000
      # ipx_interface add eth2 etherii 0x0300000000
      # ipx_interface add eth3 etherii 0x0400000000
      # ipxd



 You should then wait a moment or two and check your
 /proc/net/ipx_route file and you should see it populated with the IPX
 routes relevant to your configuration and any learned from any other
 routers in the network.


 8.1.  Do I need to configure an internal network ?

 Novell has a feature called an internal network, which it uses to
 simplify routing in situations where a host has more than one network
 device connected.  This is useful in the case of a fileserver
 connected to multiple networks as it means that only one route needs
 to be advertised to reach the server regardless of which network you
 are attempting from.
 In the case of a configuration where you are not running a fileserver
 and your machine acting only as an IPX router the question is not as
 simple to answer. It has been reported that configuring for IPX/PPP
 works `better' if you also configure an internal network.

 In any case it is easy to do, but may require a rebuild of your
 kernel.  When you are working through the kernel make config you must
 answer y when asked Full internal IPX network as illustrated:



        ...
        ...
       Full internal IPX network (CONFIG_IPX_INTERN) [N/y/?] y
        ...
        ...



 To configure the internal network interface, use the ipx_internal_net
 command described earlier in the IPX tools section. The main
 precaution to take is to ensure that they IPX network address you
 assign is unique on your network and that no other machine or network
 is using it.


 9.  Configuring your Linux machine as an NCP client.

 If you are a user of a mixed technology network that comprises both IP
 and IPX protocols it is likely that at some time or another you have
 wanted to have your Linux machine access data stored on a Novell
 fileserver on your network. Novell have long offered an NFS server
 package for their fileservers that would allow this, but if you are a
 small site or have only a small number of people interested in doing
 this it is difficult to justify the cost of the commercial package.

 Volker Lendecke <[email protected]> has written a Linux
 filesystem kernel module that supports a subset of the Novell NCP that
 will allow you to mount Novell volumes into your Linux filesystem
 without requiring any additional products for your fileserver.  Volker
 has called the package ncpfs and derived the necessary information
 mainly from the book "Netzwerkprogrammierung in C" by Manfred Hill and
 Ralf Zessin (further details of the book are contained within the
 README file in the ncpfs package).

 The software causes Linux to emulate a normal Novell workstation for
 file services. It also includes a small print utility that allows you
 to print to Novell print queues (This is documented in the Print
 Client section later).  The ncpfs package will work with Novell
 fileservers of version 3.x and later, it will not work the Novell 2.x.
 The ncpfs client will also work with close Novell compatible products,
 but unfortunately some products that claim to be compatible aren't
 compatible enough. To use ncpfs with Novell 4.x fileservers, it is
 preferred to use the Novell server in bindery emulation mode. The NDS
 support is a very recent early beta addition to ncpfs and additionally
 its use may be prohibited in your country due to the inclusion of
 patented technology.



 9.1.  Obtaining ncpfs .

 The latest ncpfs package was designed to be built against the version
 1.2.13 kernel or kernels later than 1.3.71 (this includes 2.x.x).  If
 you not using a kernel in either of these categories then you will
 have to upgrade your kernel. The Kernel-HOWTO <Kernel-HOWTO.html>
 describes how to do this in detail.

 You can obtain the ncpfs package by anonymous ftp from Volker's home
 site at: ftp.gwdg.de <ftp://ftp.gwdg.de/pub/linux/misc/ncpfs/> or
 sunsite.unc.edu
 <ftp://sunsite.unc.edu/pub/Linux/system/filesystems/ncpfs> or mirror
 sites. The current version at the time of writing was:

 ncpfs-2.0.11.tgz or ncpfs-2.2.0.tgz which adds the NDS support.


 9.2.  Building ncpfs  for kernel 1.2.13.


    Build a kernel with Ethernet and IPX support
       The first thing you need to do is ensure that your kernel has
       been built with IPX support enabled.  In the 1.2.13 version
       kernel you need only ensure that you have answered Y to the
       question: 'The IPX protocol' as illustrated:

        ...
        ...
       Assume subnets are local (CONFIG_INET_SNARL) [y]
       Disable NAGLE algorithm (normally enabled) (CONFIG_TCP_NAGLE_OFF) [n]
       The IPX protocol (CONFIG_IPX) [n] y
       *
       * SCSI support
        ...
        ...


    You will also need to ensure that you include an appropriate driver
    for your Ethernet card. If you do not know how to do this then you
    should read the Ethernet-HOWTO <Ethernet-HOWTO.html>.

    You can then proceed to build your kernel. Make sure you remember
    to run lilo to install it when you have finished.


    Untar the ncpfs software

       # cd /usr/src
       # tar xvfz ncpfs-2.0.11.tgz
       # cd ncpfs



    Check the Makefile
       If you intend to use kerneld to autoload the ncpfs kernel module
       then you must uncomment the line in the Makefile that refers to:
       KERNELD. If you are unsure what this means then you should read
       the Kernel-HOWTO <Kernel-HOWTO.html> to familiarise yourself
       with kernel module configuration.


    Make the ncpfs software
       The software should compile cleanly with no other configuration
       necessary:

       # make



    Copy the IPX tools somewhere useful if you don't already have them.
       After the make has completed you should find all of the tools
       you need in the ncpfs/bin directory. You can use:



         # make install



    to install the tools in Volkers choice of directories. If you are
    running on an ELF based system then you will need to rerun
    `ldconfig -v' to ensure that the shared library is able to be
    found.


    Copy the ncpfs.o module somewhere useful if necessary.
       If you are compiling for a 1.2.* kernel then you will find a
       file called ncpfs.o in the ncpfs/bin directory after the make
       has completed. This is the ncpfs kernel module.  You should copy
       this somewhere useful. On my debian system I have copied it to
       the /lib/modules/1.2.13/fs directory and added ncpfs to the
       /etc/modules file so that it will be automatically started at
       boot time. If you are using some other distribution you should
       find where it keeps its modules and copy it there, or just copy
       it to your /etc directory. To load the modules manually you need
       to use the command:

       # insmod ncpfs.o



 9.3.  Building ncpfs  for kernels 1.3.71++/2.0.*.

 For the latest version of ncpfs you must use kernel 1.3.71 or newer,
 this includes the 2.0.* kernels.

 If you intend using a kernel that is version 1.3.71 or newer then the
 ncpfs kernel code has been included in the standard kernel
 distribution. You need only answer Y to:


      Networking options  --->
          ...
          ...
          <*> The IPX protocol
          ...
      Filesystems  --->
          ...
          ...
          <*> NCP filesystem support (to mount NetWare volumes)
          ...



 You will still need to follow the instructions for building for
 kernels 1.2.* so that you can build the tools but there will not be a
 module file for you to install.



 9.4.  Configuring and using ncpfs .


    Configure the IPX network software
       There are two ways of configuring the IPX network software. You
       can manually configure all of your IPX network information or
       you can choose to let the software determine for itself some
       reasonable settings using the command:



         # ipx_configure --auto_interface=on --auto_primary=on



    This should be reasonable in most circumstances, but if it doesn't
    work for you then read the 'IPX tools' section above to configure
    your software manually. Problems have been noted using this on
    networks containing Windows '95 clients.


    Test the configuration
       After your IPX network is configured you should be able to use
       the slist command to see a list of all of the Novell fileserver
       on your network:

       # slist


    If the slist command displays a message like: ncp_connect: Invalid
    argument then your kernel probably does not support IPX. Check that
    you have actually booted off the appropriate kernel.  When you boot
    you should see messages about 'IPX' and 'ncpfs' in the system
    startup messages. If the slist command does not list all of your
    fileservers then you may need to use the manual network configura-
    tion method.


    Mount a Novell(tm) server or volume.
       If your IPX network software is working ok you should now be
       able to mount a Novell fileserver or volume into your Linux
       filesystem. The ncpmount command is used for this purpose and
       requires that you specify at least the following information:

       1. The fileserver name

       2. (optionally) The fileserver directory to mount

       3. The fileserver login id. If it has a password you will also
          need that.

       4. The mount point ie. where you want the mount to go. This will
          be an existing directory on your machine.

       There is an equivalent ncpumount command to unmount a mounted
       NCP filesystem. The NCP filesystems will be unmounted cleanly if
       you shutdown your machine normally, so you needn't worry about
       ncpumounting your filesystems manually before a halt or
       shutdown.

       An example command to mount fileserver ACCT_FS01, with a login
       id of guest with no password, under the /mnt/Accounts directory
       might look like the following:


       # ncpmount -S ACCT_FS01 /mnt/Accounts -U guest -n


    Note the use of the -n option to indicate that no password is
    required for the login. The same login specifying a password of
    secret would look like:

    # ncpmount -S ACCT_FS01 /mnt/Accounts -U guest -P secret


    If you don't specify either the -n or the -P options you will be
    prompted for a password.


    Check the mount
       If the mount is successful you will find all the volumes
       accessible to the userid used for login listed as directories
       under the mount point. You should then also be able to traverse
       the directory structure to find other files. You may
       alternatively use the -V option to mount a single volume.

       NCP does not provide uid or gid ownership of files. All the
       files will have the permission and ownership assigned to the
       mount point directory restricted by trustee permissions on the
       Novell server. Bear this in mind when sharing mounts between
       Linux users.


    Configure mounts to be automatically performed.
       If you have some need to permanently have an ncp mount then you
       will want to configure the commands above into your rc files so
       that they occur automatically at boot time. If your distribution
       doesn't already provide some way of configuring IPX like debian
       then I recommend you place them in your /etc/rc.local file if
       you have one. You might use something like:


         #
         # Start the ncp filesystem

         /sbin/insmod /lib/modules/1.2.13/fs/ncpfs.o

         # configure the IPX network
         ipx_configure --auto_interface=on --auto_primary=on

         # guest login to the Accounting fileserver
         ncpmount -S ACCT_FS01 /mnt/Accounts -U guest -n

         #



    There is another means of configuring NCP mounts and that is by
    building a $HOME/.nwclient file. This file contains details of tem-
    porary or user specific NCP mounts that would be performed regu-
    larly. It allows you to store the details of mounts so that you can
    recreate them without having to specify all of the detail each
    time.

    Its format is quite straightforward:



    # The first entry is the 'preferred server' entry and is
    # used whenever you do not specify a server explicitly.
    #
    # User TERRY login to DOCS_FS01 fileserver with password 'password'
    DOCS_FS01/TERRY password
    #
    # Guest login to the ACCT_FS01 fileserver with no password.
    ACCT_FS01/GUEST -



    To activate these mounts you could use:


         $ ncpmount /home/terry/docs



    to mount: DOCS_FS01 with a login of TERRY under the
    /home/terry/docs directory. Note that this entry was chosen because
    no fileserver was specified in the mount command. If the following
    command were used:


         $ ncpmount -S ACCT_FS01 /home/terry/docs



    then a GUEST login to ACCT_FS01 would be mounted there instead.

    Note: for this mechanism to work the permissions of the
    $HOME/.nwclient file must be 0600 so you would need to use the
    command:


         $ chmod 0600 $HOME/.nwclient



    If non-root users are to be allowed to use this mechanism then the
    ncpmount command must be Set Userid Root, so you would need to give
    it permissions:


         # chmod 4755 ncpmount



    Try out the nsend utility
       a utility to send messages to Novell users is also included in
       the package, it is called nsend and is used as follows:

       # nsend rod hello there


    would send the message "hello there" to a logged in user "rod" on
    your "primary" fileserver (the first one appearing in your
    .nwclient file. You can specify another fileserver with the same
    syntax as for the ncpmount command.

 10.  Configuring your Linux machine as an NCP server.

 There are two packages available that allow Linux to provide the
 functions of a Novell Fileserver. They both allow you to share files
 on your linux machine with users using Novell NetWare client software.
 Users can attach and map filesystems to appear as local drives on
 their machines just as they would to a real Novell fileserver. You may
 want to try both to see which best serves your intended purpose.


 10.1.  The mars_nwe  package.

 Martin Stover <[email protected]> developed mars_nwe to enable linux
 to provide both file and print services for NetWare clients.

 In case you are wondering about the name: mars_nwe is Martin Stovers
 Netware Emulator.


 10.1.1.  Capability of mars_nwe .

 mars_nwe implements a subset of the full Novell NCP for file services,
 disk based bindery and also print services. It is likely to contain
 bugs but there are many people using it now and the number of bugs is
 steadily decreasing as new versions are released.


 10.1.2.  Obtaining mars_nwe .

 You can obtain mars_nwe from ftp.gwdg.de
 <ftp://ftp.gwdg.de/pub/linux/misc/ncpfs/> or from
 <ftp://sunsite.unc.edu/pub/Linux/system/filesystems/ncpfs/>.

 The version current at the time of writing was:
 mars_nwe-0.99.pl10.tgz.


 10.1.3.  Building the mars_nwe  package.


    Build a kernel with Ethernet and IPX Support
       In the 1.2.13 version kernel you need only ensure that you have
       answered Y to the question: 'The IPX protocol' and N to the
       question: `Full internal IPX network' as illustrated:

        ...
        ...
       The IPX protocol (CONFIG_IPX) [n] y
        ...
        ...
       Full internal IPX network (CONFIG_IPX_INTERN) [N/y/?] n
        ...
        ...


    In newer kernels a similar process is adopted but the actual text
    of the prompt may have changed slightly.

    You will also need to ensure that you include an appropriate driver
    for your Ethernet card. If you do not know how to do this then you
    should read the Ethernet-HOWTO <Ethernet-HOWTO.html>.

    You can then proceed to build your kernel. Make sure you remember
    to run lilo to install it when you have finished.


    Untar the mars_nwe package.


         # cd /usr/src
         # tar xvfz mars_nwe-0.99.pl10.tgz



    Make mars_nwe.
       To make the package is very simple.  The first step is to simply
       run make, this will create a config.h file for you. Next you
       should look at and edit the config.h file if necessary. It
       allows you to configure items such as the installation
       directories that will be used and the maximum number of sessions
       and volumes that the server will support. The really important
       entries to look at are:

       FILENAME_NW_INI       the location of the initialisation file
       PATHNAME_PROGS        where the executable support programs will be found.
       PATHNAME_BINDERY      where the 'bindery' files will go.
       PATHNAME_PIDFILES     the directory for the 'pid' files to be written.
       MAX_CONNECTIONS       the maximum number of simultaneous connections allowed.
       MAX_NW_VOLS           the maximum number of volumes mars_nwe will support.
       MAX_FILE_HANDLES_CONN the maximum number of open files per connection.
       WITH_NAME_SPACE_CALLS if you want to support ncpfs clients.
       INTERNAL_RIP_SAP      whether you want mars_nwe to provide rip/sap routing.
       SHADOW_PWD            whether you use shadow passwords or not.



    The defaults will probably be ok but you should check anyway.

    When this is done:


         # make
         # make install



    will build the servers and install them in the appropriate
    directory. The installation script also installs the configuration
    file /etc/nwserv.conf.


    Configure the server.
       Configuration is fairly simple. You need to edit the
       /etc/nwserv.conf file. The format of this file may at first look
       a little cryptic, but it is fairly straightforward. The file
       contains a number of single line configuration items. Each line
       is whitespace delimited and begins with a number that indicates
       the contents of the line. All characters following a '#'
       character are considered a comment and ignored. Martin supplies
       an example configuration file in the package, but I'll present
       what I consider to be a simplified example to offer an
       alternative for you.



    # VOLUMES (max. 5)
    # Only the SYS volume is compulsory. The directory containing the SYS
    # volume must contain the directories: LOGIN, PUBLIC, SYSTEM, MAIL.
    # The 'i' option ignores case.
    # The 'k' option converts all filenames in NCP requests to lowercase.
    # The 'm' option marks the volume as removable (useful for cdroms etc.)
    # The 'r' option set the volume to read-only.
    # The 'o' option indicates the volume is a single mounted filesystem.
    # The 'P' option allows commands to be used as files.
    # The 'O' option allows use of the OS/2 namespace
    # The 'N' option allows use of the NFS namespace
    # The default is upper case.
    # Syntax:
    #    1 <Volumename> <Volumepath>   <Options>

    1   SYS         /home/netware/SYS/               # SYS
    1   DATA        /home/netware/DATA/      k       # DATA
    1   CDROM       /cdrom                   kmr     # CDROM

    # SERVER NAME
    # If not set then the linux hostname will be converted to upper case
    # and used. This is optional, the hostname will be used if this is not
    # configured.
    # Syntax:
    #    2 <Servername>

    2   LINUX_FS01

    # INTERNAL NETWORK ADDRESS
    # The Internal IPX Network Address is a feature that simplifies IPX routing
    # for multihomed hosts (hosts that have ports on more than one IPX network).
    # Syntax:
    #    3 <Internal Network Address> [<Node Number>]
    # or:
    #    3 auto
    #
    # If you use 'auto' then your host IP address will be used. NOTE: this may
    # be dangerous, please be sure you pick a number unique to your network.
    # Addresses are 4byte hexadecimal (the leading 0x is required).

    3   0x49a01010  1


    # NETWORK DEVICE(S)
    # This entry configures your IPX network. If you already have your
    # IPX network configured then you do not need this. This is the same as
    # using ipx_configure/ipx_interface before you start the server.
    # Syntax:
    #    4 <IPX Network Number> <device_name> <frametype> [<ticks>]
    #                         Frame types: ethernet_ii, 802.2, 802.3, SNAP

    4  0x39a01010  eth0  802.3  1

    # SAVE IPX ROUTES AFTER SERVER IS DOWNED
    # Syntax:
    #    5 <flag>
    #       0 = don't save routes, 1 = do save routes

    5 0

    # NETWARE VERSION
    # Syntax:
    #    6 <version>
    #      0 = 2.15, 1 = 3.11

    6 1
    # PASSWORD HANDLING
    # Real Novell DOS clients support a feature which encypts your
    # password when changing it. You can select whether you want your
    # mars server to support this feature or not.
    # Syntax
    #    7 <flag>
    #    <flag> is:
    #        0 to force password encryption. (Clients can't change password)
    #        1 force password encryption, allow unencrypted password change.
    #        7 allow non-encrypted password but no empty passwords.
    #        8 allow non-encrypted password including empty passwords.
    #        9 completely unencrypted passwords (doesn't work with OS/2)

    7 1

    # MINIMAL GID UID rights
    # permissions used for attachments with no login. These permissions
    # will be used for the files in your primary server attachment.
    # Syntax:
    #    10 <gid>
    #    11 <uid>
    #    <gid> <uid> are from /etc/passwd, /etc/groups

    10  200
    11  201

    # SUPERVISOR password
    # May be removed after the server is started once. The server will
    # encrypt this information into the bindery file after it is run.
    # You should avoid using the 'root' user and instead use another
    # account to administer the mars fileserver.
    #
    # This entry is read and encrypted into the server bindery files, so
    # it only needs to exist the first time you start the server to ensure
    # that the password isn't stolen.
    #
    # Syntax:
    #    12 <Supervisor-Login> <Unix username> [<password>]

    12  SUPERVISOR  terry  secret

    # USER ACCOUNTS
    # This associates NetWare logins with unix accounts. Password are
    # optional.
    # Syntax:
         13 <User Login> <Unix Username> [<password>]

    13  MARTIN martin
    13  TERRY  terry

    # LAZY SYSTEM ADMIN CONFIGURATION
    # If you have a large numbers of users and could not be bothered using
    # type 13 individual user mappings, you can automatically map mars_nwe
    # logins to linux user names. BUT, there is currently no means of making
    # use of the linux login password so all users configured this way are
    # will use the single password supplied here. My recommendation is not
    # to do this unless security is absolutely no concern to you.
    # Syntax:
    #    15 <flag> <common-password>
    #    <flag> is: 0  - don't automatically map users.
    #               1  - do automatically map users not configured above.
    #               99 - automatically map every user in this way.

    15  0  duzzenmatta

    # SANITY CHECKING
    # mars_nwe will automatically ensure that certain directories exist if
    # you set this flag.
    # Syntax:
    #    16 <flag>
    #    <flag> is 0 for no, don't, or 1 for yes, do.

    16  0

    # PRINT QUEUES
    # This associates NetWare printers with unix printers. The queue
    # directories must be created manually before printing is attempted.
    # The queue directories are NOT lpd queues.
    # Syntax:
    #    21 <queue_name> <queue_directory> <unix_print_cmd>

    21  EPSON  SYS:/PRINT/EPSON lpr -h
    21  LASER  SYS:/PRINT/LASER lpr -Plaser

    # DEBUG FLAGS
    # These are not normally needed, but may be useful if are you debugging
    # a problem.
    # Syntax:
    #    <debug_item> <debug_flag>
    #
    #    100 = IPX KERNEL
    #    101 = NWSERV
    #    102 = NCPSERV
    #    103 = NWCONN
    #    104 = start NWCLIENT
    #    105 = NWBIND
    #    106 = NWROUTED
    #                 0 = disable debug, 1 = enable debug

    100 0
    101 0
    102 0
    103 0
    104 0
    105 0
    106 0

    # RUN NWSERV IN BACKGROUND AND USE LOGFILE
    # Syntax:
    #    200 <flag>
    #        0 = run NWSERV in foreground and don't use logfile
    #        1 = run NWSERV in background and use logfile

    200  1

    # LOGFILE NAME
    # Syntax:
    #    201 <logfile>

    201  /tmp/nw.log

    # APPEND LOG OR OVERWRITE
    # Syntax:
    #    202 <flag>
    #        0 = append to existing logfile
    #        1 = overwrite existing logfile

    202  1

    # SERVER DOWN TIME
    # This item sets the time after a SERVER DOWN is issued that the
    # server really goes down.
    # Syntax:
    #    210 <time>
    #        in seconds. (defaults 10)

    210  10

    # ROUTING BROADCAST INTERVAL
    # The time is seconds between server broadcasts
    # Syntax:
    #    211 <time>
    #        in seconds. (defaults 60)

    211  60

    # ROUTING LOGGING INTERVAL
    # Set how many broadcasts take place before logging of routing
    # information occurs.
    # Syntax:
    #    300  <number>

    300  5

    # ROUTING LOGFILE
    # Set the name of the routing logfile
    # Syntax:
    #    301 <filename>

    301  /tmp/nw.routes

    # ROUTING APPEND/OVERWRITE
    # Set whether you want to append to an existing log file or
    # overwrite it.
    # Syntax:
    #    302 <flag>
    #        <flag> is 0 for append, 1 for create/overwrite

    302  1

    # WATCHDOG TIMING
    # Set the timing for watchdog messages that ensure the network is
    # still alive.
    # Syntax:
    #    310 <value>
    #        <value> =   0 - always send watchdogs
    #                  < 0 - (-ve) for disable watchdogs
    #                  > 0 - send watchdogs when network traffic
    #                        drops below 'n' ticks

    310  7

    # STATION FILE
    # Set the filename for the stations file which determine which
    # machines this fileserver will act as the primary fileserver for.
    # The syntax of this file is described in the 'examples' directory
    # of the source code.
    # Syntax:
    #    400 <filename>

    400  /etc/nwserv.stations

    # GET NEAREST FILESERVER HANDLING
    # Set how SAP Get Nearest Fileserver Requests are handled.
    # Syntax:
    #    401 <flag>
    #        <flag> is: 0 - disable 'Get Nearest Fileserver' requests.
    #                   1 - The 'stations' file lists stations to be excluded.
    #                   2 - The 'stations' file lists stations to be included.

    401  2



    Start the server
       If you've configured the server to expect external programs to
       configure your network and/or provide the routing function then
       you should start those before starting the server. Presuming you
       have configured the server so that it will configure your
       interfaces for you and provide the routing services you need
       only issue the command:


         # nwserv



    Test the server
       To test the server you should first try to attach and login from
       a NetWare client on your network. You then set a CAPTURE from
       the client and attempt a print. If both of these are successful
       then the server is working.


 10.2.  The lwared  package.

 Ales Dryak <[email protected]> developed lwared to allow Linux to
 function as an NCP based fileserver.

 Ales has called the package lwared, an abbreviation for LinWare
 Daemon.


 10.2.1.  Capability of lwared .

 The lwared server is capable of providing a subset of the full
 function of the Novell NCP. It incorporates messaging but it does not
 provide any printing facilities at all. It does not currently work
 very well with either Windows95 or Windows NT clients. The lwared
 server relies on external programs to build and update the IPX routing
 and SAP tables.  Misbehaving clients can cause the server to crash.
 Importantly, filename translation facilities have not been included.

 The server does work for NETX and VLM NetWare shells.


 10.2.2.  Obtaining lwared

 The lwared package can be built for any kernel newer than 1.2.0, I
 recommend you use version 1.2.13 as no kernel patches are required if
 you do. Some of the IPX functionality has changed with the version
 1.3.* kernels and this means that patches are now required to make it
 work properly. Appropriate patches are included for the new kernels,
 so if you must use an alpha kernel you should still be able to get
 lwared to work properly for you.

 You can obtain the lwared package by anonymous ftp from:
 klokan.sh.cvut.cz <ftp://klokan.sh.cvut.cz/pub/linux/linware/>


 or from:

 sunsite.unc.edu
 <ftp://sunsite.unc.edu/pub/Linux/system/network/daemons> or mirror
 sites. The current version at the time of writing was:
 lwared-0.95.tar.gz


 10.2.3.  Building lwared


    Untar the lwaredpackage
       Something like:

       # cd /usr/src
       # tar xvpfz lwared-0.95.tar.gz



    Build a kernel with Ethernet and IPX support
       If you are using an alpha 1.3.* kernel then you should try and
       use kernel version 1.3.17 or newer because the supplied patches
       were built against it. 1.3.* kernels older than 1.3.17 will
       require hand patching to install. (some information on how to do
       this is included in the INSTALL file in the package.). To
       install the patches against a 1.3.17 kernel or newer you should
       try:

       # make patch



    After applying the patches if necessary, the next thing you need to
    do is ensure that your kernel has been built with IPX support
    enabled. In the 1.2.13 version kernel you need only ensure that you
    have answered Y to the question: 'The IPX protocol' as illustrated:

     ...
     ...
    Assume subnets are local (CONFIG_INET_SNARL) [y]
    Disable NAGLE algorithm (normally enabled) (CONFIG_TCP_NAGLE_OFF) [n]
    The IPX protocol (CONFIG_IPX) [n] y
    *
    * SCSI support
     ...
     ...


    In newer kernels a similar process is adopted by the actual text of
    the prompt may have changed slightly.

    You will also need to ensure that you include an appropriate driver
    for your Ethernet card. If you do not know how to do this then you
    should read the Ethernet-HOWTO <Ethernet-HOWTO.html>.

    You can then proceed to build your kernel. Make sure you remember
    to run lilo to install it when you have finished.


    Compile and install lwared.
       To compile lwared you should first check, edit if necessary, the
       server/config.h file. This file contains various settings that
       will govern the way your server will behave when it is running.
       The defaults are reasonable, though you might want to check that
       the directories specified for the log files and configuration
       files suit your system.

       # make depend
       # make
       # make install


    I found that the 'make depend' complained about not finding the
    float.h file on my system but appeared to work anyway.  I also
    found that when I tried compiling with gcc 2.6.3 I found I had to
    change the line:

    #include <net/route.h>


    to

    #include <net/if_route.h>


    in lib/ipxkern.c as this file changed name sometime.

    The 'make install' will attempt to install the server and routing
    daemon programs into your /usr/sbin directory, the lwpasswd program
    into your /usr/bin directory, the IPX utility programs will be
    installed into your /sbin directory and last but not least the
    manual pages will go into the /usr/man directory structure. If any
    of these locations are not suitable for your system then you should
    edit the relevant Makefile and change the target directories to
    suit.



 10.2.4.  Configuring and using lwared

 Now the fun bit!


    Configuring the IPX network
       The first thing you must do is configure your Ethernet
       interfaces to support the IPX networks your server will support.
       To do this you will need to know the IPX network addresses for
       each of your LAN segments, which Ethernet device (eth0, eth1
       etc.) is on which segment, what frame type (802.3, EtherII etc.)
       each LAN segment uses and what Internal Network address your
       server should use (this is really needed if your server will
       service more than one LAN segment).  A configuration for a
       server that is on two dis-similar segments with IPX network
       addresses 23a91300 and 23a91301 and internal network address
       bdefaced might look like:

       # ipx_internal_net add BDEFACED 1
       # ipx_interface add eth0 802.3 23a91300
       # ipx_interface add eth1 etherii 23a91301



    Start the routing daemons
       The kernel software itself actually does the IPX packet
       forwarding as it does for IP, but the kernel requires additional
       programs to manage the routing table updates. In the case of IPX
       two daemons are needed and both are supplied with lwared:
       ipxripd manages the IPX routing information and ipxsapd manages
       the SAP information.  To start the daemons you need only specify
       the location of where they should write their log messages:
       # ipxripd /var/adm/ipxrip
       # ipxsapd /var/adm/ipxsap



    Configure the lwared server
       There are two files that you must manually configure to allow
       user login to your lwared server.  They are:



       /etc/lwpasswd
          This is where LinWare user account information is kept. The
          lwpasswd program is to keep it up to date.  In its simplest
          form the /etc/lwpasswd file looks like:

          ales:
          terryd:
          guest:


       Its format is a simple list of login id followed by a ':' char-
       acter and then the encrypted version of the login passwd. A cou-
       ple of important caveats here: No encrypted password means no
       password, LinWare users must have Linux accounts, that is any
       user you place in /etc/lwpasswd must also appear in /etc/passwd
       and root is the only account that can change the password of
       another LinWare user. If you are logged in as root you can
       change the password of a LinWare user as this transcript demon-
       strates:

       # lwpasswd rodg
       Changing password for RODG
       Enter new password:
       Re-type new password:
       Password changed.



       /etc/lwvtab
          This is the LinWare volume tables and it stores information
          about what directories should be made available to LinWare
          users (this file is similar in nature to the NFS /etc/exports
          file). A simple example of its format is as follows:

          SYS             /lwfs/sys
          DATA            /lwfs/data
          HOME            /home


       The format is simple: Volume name followed by whitespace fol-
       lowed by Linux directory to export. You must have at least an
       entry for the SYS volume for the server to start. If you intend
       your DOS based users to be able use your LinWare server as their
       primary server then you must install a standard SYS volume
       directory structure underneath the directory you export as your
       SYS volume. Since these files are proprietary and copyright to
       the Novell corporation you should have a license for these. If
       you users will be using a Novell fileserver as their primary
       server then this will not be necessary.


    Start the lwared server.
       tada!

       # lwared


    It is almost an anticlimax isn't it ? Ok so you've got a question,
    right?  What is the fileserver name that is being advertised ? If
    you started the server as shown then the LinWare server name being
    advertised will be based on what is returned by the Linux hostname.
    If you'd like it to be something else then you can give the server
    the name when you start it, for example:

    # lwared -nlinux00


    would start the server with the name linux00.


    Test the lwared server.
       The very first thing to test is that your LinWare server appears
       in an slist from a DOS client on your network. The slist program
       is stored on the SYS volume of a Novell fileserver so you must
       do this from a machine that is already logged in somewhere. If
       this is not successful then check that ipxsapd and lwared are
       both running. If the slist is successful then you should try
       attaching to the server and mapping a volume:

       C:> attach linux00/ales
        ...
        ...
       C:> map l:=linux00/data:
       C:> l:


    You should then be able to treat the new map just like any other
    map. The file permissions you will have will be based on those
    allowed to the linux account that parallels your LinWare login.



 11.  Configuring your Linux machine as a Novell Print Client.

 The ncpfs package includes two small programs that allow you to handle
 printing from you Linux machine to a printer attached to a Novell
 print server.  The nprint command allows you to print to a file to a
 NetWare print queue. The pqlist command allows you the list the
 available print queues on a NetWare server.

 To obtain and install these commands just follow the instructions
 relating to the NCP client described earlier.

 Both commands require that you supply username and password so you
 might normally consider building some shell scripts to make the task
 of printing easier.

 An example might look like:


      # pqlist -S ACCT_FS01 -U guest -n
      # nprint -S ACCT_FS01 -q LASER -U guest -n filename.txt



 The login syntax is similar to the ncpmount command. The examples
 above assume that fileserver ACCT_FS01 has a guest account with no
 password, that a print queue called LASER exists and that guest is
 allowed to print to it.
 On my Linux boxen I have a short shell script for each Novell printer.
 This can then be used as a print filter to allow printing using the
 standard Linux spooler.


 12.  Configuring your Linux machine as a Novell Print Server.

 A program to allow your Linux machine to act as a print server on a
 Netware network is included in the ncpfs package. For instructions on
 how to obtain and build, it follow the directions in the `Netware
 client' section above. Alternatively, support is included in the
 mars_nwe package.



 12.1.  Prerequisites

 Configuration is quite straightforward but relies on you already
 having your printer configuration completed and working under Linux.
 This is covered in the Printing-HOWTO <Printing-HOWTO.html> in some
 depth.


 12.2.  Configuration

 When you have a working printer configuration, and you have built and
 installed the pserver utility then you need to add commands to start
 it into your rc files.

 Exactly what command will use will depend on depend on exactly how you
 want it to operate, but in its simplest form something like the
 following will work:



      # pserver -S ACCT_01 -U LASER -P secret -q LASERJET



 This example asks the pserver utility to login in to the ACCT_01
 fileserver with username LASER and password secret and to take jobs
 from the LASERJET print queue. When an incoming print job is received
 it will use the default print command of lpr to feed the print job to
 the Linux print daemon. The print queue must already be defined on the
 fileserver and the username must have server priveliges for the queue.

 You could if you wished use any Linux command to accept and print the
 print job. The -c argument allows you to specify the exact print
 command.  For example:


      # pserver -S ACCT_01 -U LASER -P secret -q LASERJET -c "lpr -Plaserjet"



 would do exactly the same as the previous example except it would send
 the job to the laserjet printcap configuration instead of the default
 one.


 13.  An overview of the ncpfs  user and adminstration commands

 Recent versions of Volker's ncpfs package include a range of user and
 administration commands that you might want to use. The tools are
 built and installed as part of the ncpfs installation process, so if
 you haven't already, follow the instructions supplied in the Novell
 Client section above to build and install them.

 Detailed information is available in the supplied man pages but a
 brief summary of the commands is as follows;


 13.1.  User commands.


    ncopy
       Network Copy - allows efficient file copies to be performed by
       using a Netware function rather than a copy across the network.

    nprint
       Network Print - allows you to print a file to a Netware print
       queue on a Netware server.

    nsend
       Network Send - allows you to send messages to other users on a
       Netware server.

    nwbols
       List Bindery Objects - allows you to list the bindery contents
       of a Netware server.

    nwboprops
       List Properties of a Bindery Object - allows you to the
       properties of a Netware bindery object.

    nwbpset
       Set Bindery Property - allows you to set the properties of a
       Netware bindery object.

    nwbpvalues
       Print Netware Bindery Objects Property Contents - allows you to
       print the contents of a Netware bindery property.

    nwfsinfo
       Fileserver Information - prints some summary information about a
       Netware server.

    nwpasswd
       Netware Password - allows you to change a Netware users
       password.

    nwrights
       Netware Rights - displays the rights associated with a
       particular file or directory.

    nwuserlist
       Userlist - lists the users currently logged into a Netware
       fileserver.

    pqlist
       Print Queue List - displays the contents of a Netware print
       queue.

    slist
       Server List - displays a list of know Netware fileserver.



 13.2.  Administration tools.


    nwbocreate
       Create a Bindery Object - allows you to create a Netware bindery
       object.

    nwborm
       Remove Bindery Object - allows you to delete a Netware bindery
       object.

    nwbpadd
       Add Bindery Property - allows you to set the value of an
       existing property of a Netware bindery object.

    nwbpcreate
       Create Bindery Property - allows you to create a new property
       for an existing Netware bindery object.

    nwbprm
       Remove Bindery Property - allows you to remove a property from a
       Netware bindery object.

    nwgrant
       Grant Trustee Rights - allows you to assign trustee rights to a
       directory on a Netware fileserver.

    nwrevoke
       Revoke Trustee Rights - allows you to remove trustee rights from
       a directory on a Netware fileserver.



 14.  Configuring PPP for IPX support.

 New versions of the pppd PPP daemon for Linux have support that allows
 you to carry IPX packets across a PPP serial link. You need at least
 version ppp-2.2.0d of the daemon. See the PPP-HOWTO <PPP-HOWTO.html>
 for details on where to find it. When you compile pppd you must ensure
 you enable the IPX support by adding the following two lines:



      IPX_CHANGE = 1
      USE_MS_DNS = 1



 to: /usr/src/linux/pppd-2.2.0f/pppd/Makefile.linux.

 The IPX_CHANGE is what configures the IPX support into PPP.  The
 USE_MS_DNS define allows Microsoft Windows95 machines to do Name
 Lookups.

 The real trick to getting it to work in knowing how to configure it.

 There are many ways of doing this, but I'm only going to describe the
 two that I've received any information on. I've tried neither yet, so
 consider this section experimental, and if you get something to work,
 please let me know.



 14.1.  Configuring an IPX/PPP server.

 The first thing you need to do is configure your Linux machine as an
 IP/PPP server. Don't panic! This isn't difficult. Again, follow the
 instructions in the PPP-HOWTO <PPP-HOWTO.html> and you should be
 pretty much ok. When you have this done there are a couple of simple
 modifications you need to make to get IPX working over the same
 configuration.


 14.1.1.  First steps.

 One of the first steps you must take is to configure your linux
 machine as an IPX router as described in the appropriate section
 earlier in this document.  You won't need to use the ipx_route command
 for the ppp interface because pppd will configure these for you as it
 does for IP. When you have the ipxd daemon running it will
 automatically detect any new IPX interfaces and propogates routes for
 them. In this way your dialup hosts will be seen by other machines
 automatically when they connect.



 14.1.2.  Design.

 When you are running as a server it will normally be your
 responsibility to assign network address to each of the PPP links when
 they are established.  This is an important point, each PPP link will
 be an IPX network and will have a unique IPX network address. This
 means that you must decide how you will allocate addresses and what
 what they will be. A simple convention is to allocate one IPX network
 address to each serial device that will support IPX/PPP. You could
 allocate IPX network addresses based on the login id of the connecting
 user, but I don't see any particularly good reason to do so.

 I will assume that this is what you have done, and that there are two
 serial devices (modems) that we will use. The addresses I've assigned
 in this contrived example are:



      device IPX Network Address
      ------ -------------------
      ttyS0  0xABCDEF00
      ttyS1  0xABCDEF01



 14.1.3.  Configure pppd .

 Configure your /etc/ppp/options.ttyS0 file as follows:



      ipx-network 0xABCDEF00
      ipx-node 2:0
      ipxcp-accept-remote



 and your /etc/ppp/options.ttyS1 file as:


      ipx-network 0xABCDEF01
      ipx-node 3:0
      ipxcp-accept-remote



 These will ask pppd to allocate the appropriate IPX network addresses
 to the link when the link is established, set the local node number to
 2 or 3 and will let the remote node overwrite what the remote node
 number with what it thinks it is. Note that each of the addresses are
 hexadecimal numbers and that 0x is required at the start of the
 network address, but not required at the start of the node address.

 There are other places this information could be configured. If you
 have only one dialin modem then an entry could go into the
 /etc/ppp/options file. Alternatively this information can be passed on
 the command line to pppd.


 14.1.4.  Test the server configuration.

 To test the configuration you will need to have a client configuration
 that is known to work. When the caller dials in, logs in and pppd
 starts it will assign the network address, advise the client of the
 servers node number and negotiate the clients node number. When this
 has completed, and after ipxd has detected the new interface the
 client should be able to establish IPX connections to remote hosts.


 14.2.  Configuring an IPX/PPP client.

 In a client configuration, whether or not you configure your Linux
 machine as an IPX router depends on whether you have a local LAN that
 you wish to act as an IPX router for. If you are a standalone machine
 connecting to an IPX/PPP dialin server then you won't need to run
 ipxd, but if you have a LAN and wish all of the machines on the LAN to
 make use of the IPX/PPP route then you must configure and run ipxd as
 described.  This configuration is much simpler because you do not have
 multiple serial devices to configure.


 14.2.1.  Configuring pppd

 The simplest configuration is one that allows the server to supply all
 of the IPX network configuration information. This configuration would
 be compatible with the server configuration described above.

 Again you need to add some options to your /etc/ppp/options file, they
 are:


      ipxcp-accept-network
      ipxcp-accept-remote
      ipxcp-accept-local



 These options tell pppd to act completely passively and accept all of
 the configuration details from the server. You could supply default
 values here for servers that don't supply details by adding ipx-
 network and ipx-node entries similar to the server configuration.



 14.2.2.  Testing the IPX/PPP client.

 To test the client you will need a known working server to dial into.
 After you have dialled in and pppd has run you should see the IPX
 details configured on your ppp0 device when you run the ifconfig
 command and you should be able to use ncpmount.

 I'm not sure whether you will have to manually add IPX routes so that
 you can reach distant fileserver or not. This seems likely. If anyone
 running this configuration could tell me I'd be grateful.


 15.  IPX tunnel over IP

 Many of you will be in a situation where you have two Novell Local
 Area Netorks with only an IP connection between them. How do you play
 multiplayer deathmatch DOOM for DOS via this arrangement you might ask
 ? Andreas Godzina <[email protected]> has an answer for you in the form
 of ipxtunnel.

 ipxtunnel provides a bridge-like facility for IPX by allowing IPX
 packets to be encapsulated with tcp/ip datagrams so that they can be
 carried by a tcp/ip connection. It listens for IPX packets and when it
 hears one it wraps it within a tcp/ip datagram and routes it to a
 remote IP address that you specify. For this to work of course the
 machine that you route the encapsulated IPX must also be running a
 copy of the same version of ipxtunnel as you.


 15.1.  Obtaining ipxtunnel

 You can obtain ipxtunnel from sunsite.unc.edu
 <ftp://sunsite.unc.edu/pub/Linux/system/network/daemons> or mirror
 sites.


 15.2.  Building ipxtunnel

 ipxtunnel built cleanly for me using the following commands:


      # cd /usr/src
      # tar xvfz .../ipxtunnel.tgz
      # cd ipxtunnel
      # make



 15.3.  Configuring ipxtunnel

 Configuration for ipxtunnel is easy. Lets say that your friends
 machine is gau.somewhere.com and your machine is called gim.sw.edu.
 ipxtunnel uses a configuration file called /etc/ipxtunnel.conf. This
 file allows you to specify the default UDP port to use for the tcp/ip
 connection, where to send the encapsulated data and which of your
 local interfaces ipxtunnel should listen on and deliver IPX packets
 to.

 A simple configuration file would look like the following:



 #
 # /etc/ipxtunnel.conf for gim.sw.edu
 #
 # The UDP port to use:                       (default 7666)
 port 7777
 #
 # The remote machine to send IPX packets to: (no default)
 remote gau.somewhere.com
 #
 # The local interfaces to listen for IPX on: (default eth0)
 interface eth0
 interface eth1



 Obviously the other machine would have a similar configuration file
 specifying this machine as a remote host.


 15.4.  Testing and using ipxtunnel

 ipxtunnel acts like an IPX bridge, so the IPX networks at either end
 of the link should probably be the same. Andreas has never tested the
 ipxtunnel in an environment that actually supports Novell file servers
 so if you do try this in a real environment let Andreas know if it
 works or not.

 If the ipxtunnel is working you should be able to start your DOOM
 machines up at each end of the link running IPX mode and they should
 see each other.

 Andreas has only used this code over good high speed lines and he
 makes no claim as to its performance when your link is low speed.
 Again, let him know what works for you and what doesn't.



 16.  Commercial IPX support for Linux.



 16.1.  Caldera'a Network Desktop

 Caldera Inc., produce a Linux distribution that features a range of
 commercially supported enhancements including fully functional Novell
 NetWare client support. The base distribution is the well respected
 Red Hat Linux Distribution and Caldera have added their "Network
 Desktop" products to this.  The NetWare support provides a fully
 featured Novell NetWare client built on technology licensed from
 Novell Corporation. The client provides full client access to Novell
 3.x and 4.x fileservers and includes features such as NetWare
 Directory Service (NDS) and RSA encryption.

 You can obtain much more information and ordering details from the:
 Caldera Inc Web Server <http://www.caldera.com/>.

 If you work within a Netware 4.x and/or NDS environment then the
 Caldera Netware Client is the only solution available.

 If you have a business critical application for Novell support for
 Linux then the Caldera product should be something you take a close
 look at.


 17.  Some Frequently Asked Questions


    Where can I find commercially supported IPX software for Linux ?
       The Caldera Corporation offers a fully licensed and fully
       supported Netware 3.x and 4.x client. You can obtain information
       about it from the Caldera Inc Web Server
       <http://www.caldera.com/>.


    Does the IPX software work with Arcnet/Token Ring/etc. ?
       The Linux IPX software does work with ArcNet and Token Ring
       interfaces.  I haven't heard of anyone trying it with AX.25 yet.
       Configuration is the same as for configuring for ethernet except
       you will have to substitute appropriate device names in place of
       'eth0' and appopriate hardware addresses where necessary.


    How do I configure more than one IPX interface ?
       If you have more than one interface in your machine you should
       use the ipx_interface command to manually configure each one,
       you should not use the `plug n play' configuration.


    How do I choose IPX addresses ?
       IPX networking is similar, but not identical to, IP networking.
       A major difference is the way that addresses are used. IPX does
       not use the concept of subnetworking and so the sort of
       associations that you have between network addresses and
       networks is different. The rules are fairly simple:

    o  Every IPX network address must be unique on a wide area network.
       This includes Internal Network Addresses. Many organisations
       using IPX over a wide area network will have some sort of
       addressing standard that you should follow.

    o  Every Host address on an individual network must be unique. This
       means that every host on each IPX network must have a uniquely
       assigned address. In the case of ethernet network this isn't
       difficult as the cards each have a unique address. In the case
       of IPX/PPP this means you must ensure that you allocate unique
       addresses to all hosts on the network, irrespective of which end
       of the link(s) they are connected. Host address do not need to
       be unique across a wide area network as the network address is
       used in combination with the host address to uniquely identify a
       host.


    What are frame types, which should I use ?
       There are a variety of frame types in use over which you can run
       IPX.  The most common of these are described in the 'common
       terms' section of this document (under the `Frame Type entry').

       If you are installing your machine on an existing network then
       you must use whatever is already in use to allow you to
       interwork with the other hosts on the network, but if the
       installation is a brand new network you can use any of a range
       of protocols to carry  your IPX traffic. My recommendation if
       you are configuring a brand new network and you need to carry
       both IPX and IP traffic is to use the Ethernet_II frame type.


    My Windows95 machines mess up my frame type autodetection ?
       Apparently they can, yeah. I could make nasty comments, but
       instead I'll just suggest that you use the manual frame type
       configuration instead of the automatic one. It is probably the
       better way anyway.


    Why do I get the message `invalid argument' when I configure IPX ?
       You are probably not running a kernel that supports IPX, either
       recompile your kernel so it does, or double check that you have
       actually used lilo to install and run the new kernel.


    Why do I get the message `package not installed' when I configure
       IPX ?
       You are probably not running a kernel that supports IPX, either
       recompile your kernel so it does, or double check that you have
       actually used lilo to install and run the new kernel.


    Why do I get the message `IPX support not in kernel' from pppd ?
       You've probably compiled IPX as a module and not ensured that it
       was loaded before started pppd.


    How do I NFS export a mounted NCP filesystem ?
       To use NFS to export an NCP filesystem you must mount it using
       the ncpmount -V option. This option allows you to mount only one
       volume of a fileserver instead of the usual mounting of all of
       them.  When you do this your NFS daemon will allow you to export
       that filesystem in the usual way.


    Why doesn't slist work when I have an internel network with
       mars_nwe ?
       You must have the get nearest server enabled.  That is, entry
       401 in /etc/nwserv.conf should be 0 unless you have a reason for
       not responding to get nearest servers.  If you just want slist
       to work and not respond to every get nearest server request,
       include your internal network and node number in
       /etc/nwserv.stations and set entry 401 in /etc/nwserv.conf to 2.


    Does ncpfs package work with mars_nwe ?
       Martin and Volker's code is slowly beginning to converge. Recent
       versions of mars_nwe have an option to enable it to work with
       ncpfs.  You must enable the WITH_NAME_SPACE_CALLS in the
       mars_nwe config.h file.


    Is there any free DOS software to work with mars_nwe ?
       A contrived question deserves a contrived answer. I'm glad you
       asked, Martin has a package that he distributes alongside his
       mars_nwe package that offers free DOS client support for the
       mars_nwe server.  You can find it at the same sites as the
       server, and it will be called mars_dosutils-0.01.tgz. It
       includes C source code for programs such as slist.exe,
       login.exe, map.exe etc. The source is compilable with
       Borland(tm) C.



 18.  Copyright Message.

 The IPX-HOWTO, a guide to software supporting the IPX protocol for
 Linux.  Copyright (c) 1995 Terry Dawson.

 This program is free software; you can redistribute it and/or modify
 it under the terms of the GNU General Public License as published by
 the Free Software Foundation; either version 2 of the License, or (at
 your option) any later version.

 This program is distributed in the hope that it will be useful, but
 WITHOUT ANY WARRANTY; without even the implied warranty of
 MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU
 General Public License for more details.

 You should have received a copy of the GNU General Public License
 along with this program; if not, write to the:

 Free Software Foundation, Inc., 675 Mass Ave, Cambridge, MA 02139,
 USA.


 19.  Miscellaneous and Acknowledgements.

 Terry Dawson <[email protected]> for the original
 document

 David E. Storey <[email protected]> and Volker Lendecke
 <[email protected]> both assisted greatly by supplying me with
 information for this document.  Gilbert Callaghan
 <[email protected]>, David Higgins <[email protected]> and Chad
 Robinson <[email protected]> each contributed information on
 configuring IPX/PPP.  Bennie Venter <[email protected]>
 contributed some useful information relating to frame types.
 Christopher Wall <[email protected] contributed some useful suggestions
 to improve the readability and layout of the document.  Axel Boldt
 <[email protected]> contributed some useful suggestions and
 feedback.  Erik D. Olson <[email protected]> provided some useful feedback
 and information on configuring PPP for IPX.  Brian King
 <[email protected]> contributed a question for the FAQ
 section.

 "NetWare" is a registered trademark of the Novell Corporation
 <http://www.novell.com/>.  "Caldera" is a registered trademark of the
 Caldera Corporation <http://www.caldera.com/>.



 regards Kevin Thorpe.

 <[email protected]>