TCP/IP Release Document
Software Version 3.0
Part No. 005498
Revision 05
This document describes TCP/IP
Software Version 3.0.
The release notes for standard DOMAIN
Software and other optional products
are documented in other sets of
release notes and are located in the
system /doc directory.
APOLLO COMPUTER INC.
330 Billerica Road
Chelmsford, Massachusetts 01824
Copyright c 1986 Apollo Computer Inc.
All rights reserved. Printed in U.S.A.
Print Date: January, 1987.
This document was formatted using the FMT tool distributed with the DOMAIN
computer system.
APOLLO and DOMAIN are registered trademarks of Apollo Computer Inc.
AEGIS, DGR, DOMAIN/BRIDGE, DOMAIN/DFL-100, DOMAIN/DQC-100, DOMAIN/Dialogue,
DOMAIN/IX, DOMAIN/Laser-26, DOMAIN/PCI, DOMAIN/SNA, D3M, DPSS, DSEE, GMR, and
GPR are trademarks of Apollo Computer Inc.
MULTIBUS is a trademark of the Intel Corporation.
ETHERNET is a registered trademark of the Xerox Corporation.
Apollo Computer Inc. reserves the right to make changes in specifications and
other information contained in this publication without prior notice, and the
reader should in all cases consult Apollo Computer Inc. to determine whether
any such changes have been made.
THE TERMS AND CONDITIONS GOVERNING THE SALE OF APOLLO COMPUTER INC. HARDWARE
PRODUCTS AND THE LICENSING OF APOLLO COMPUTER INC. SOFTWARE CONSIST SOLELY OF
THOSE SET FORTH IN THE WRITTEN CONTRACTS BETWEEN APOLLO COMPUTER INC. AND ITS
CUSTOMERS. NO REPRESENTATION OR OTHER AFFIRMATION OF FACT CONTAINED IN THIS
PUBLICATION, INCLUDING BUT NOT LIMITED TO STATEMENTS REGARDING CAPACITY,
RESPONSE-TIME PERFORMANCE, SUITABILITY FOR USE OR PERFORMANCE OF PRODUCTS
DESCRIBED HEREIN SHALL BE DEEMED TO BE A WARRANTY BY APOLLO COMPUTER INC. FOR
ANY PURPOSE, OR GIVE RISE TO ANY LIABILITY BY APOLLO COMPUTER INC.
WHATSOEVER.
IN NO EVENT SHALL APOLLO COMPUTER INC. BE LIABLE FOR ANY INCIDENTAL,
INDIRECT, SPECIAL OR CONSEQUENTIAL DAMAGES WHATSOEVER (INCLUDING BUT NOT
LIMITED TO LOST PROFITS) ARISING OUT OF OR RELATING TO THIS PUBLICATION OR
THE INFORMATION CONTAINED IN IT, EVEN IF APOLLO COMPUTER INC. HAS BEEN
ADVISED, KNEW OR SHOULD HAVE KNOWN OF THE POSSIBILITY OF SUCH DAMAGES.
THE SOFTWARE PROGRAMS DESCRIBED IN THIS DOCUMENT ARE CONFIDENTIAL INFORMATION
AND PROPRIETARY PRODUCTS OF APOLLO COMPUTER INC. OR ITS LICENSORS.
Reader_Notice
This document resides on line in the /doc directory. To print a copy of this
document, use the prf command with the -npag and -pr options.
$ prf <filepathname> -pr <printername> -npag
iii
Contents
Chapter Page
CHAPTER 1 OVERVIEW OF TCP/IP SOFTWARE VERSION 3.0
1.1 The Subnet Utility . . . . . . . . . . . . . . . . . . 1-1
1.2 Changes to tcpstat . . . . . . . . . . . . . . . . . . 1-7
1.3 Changes to Running tcpserver with the Debug Option . 1-8
1.4 Changes to TCP/IP Installation Procedure . . . . . . . 1-10
1.5 Larger UDP Packet Size . . . . . . . . . . . . . . . . 1-10
1.6 New telnet debug Command . . . . . . . . . . . . . . . . 1-10
1.7 Additional ftp Commands . . . . . . . . . . . . . . . . 1-11
CHAPTER 2 SOFTWARE INSTALLATION PROCEDURES
2.1 Conventions and Terminology . . . . . . . . . . . . . . 2-1
2.2 Planning the Installation . . . . . . . . . . . . . . 2-2
2.3 More Information About Administrative-Mode Installations 2-3
2.4 Installation Procedures . . . . . . . . . . . . . . . . 2-4
2.5 Installed TCP/IP Files . . . . . . . . . . . . . . . . 2-15
CHAPTER 3 CHANGES IN DOCUMENTATION
CHAPTER 4 BUGS AND BUG FIXES
4.1 Bugs in Release 3.0. . . . . . . . . . . . . . . . . . . 4-1
4.2 Restrictions in TCP/IP Software . . . . . . . . . . . . 4-2
4.3 Bug Fixes Since Release 2.1 . . . . . . . . . . . . . . . 4-3
Contents iv
CHAPTER 1
OVERVIEW OF TCP/IP SOFTWARE VERSION 3.0
The DOMAIN Transmission Control Protocol/Internet Protocol (TCP/IP) product
provides file transfer and remote log-in capabilities between DOMAIN nodes
and other systems via an ETHERNET gateway.
TCP/IP Version 3.0 contains the following:
o A new feature, the subnet utility
o Changes to the tcpstat command
o Changes to running tcpserver with the -debug option
o Changes to TCP/IP installation procedure: TCP/IP No Longer Contains
Driver Software
o Larger UDP packet size
o New DOMAIN telnet debug command
o Additional DOMAIN ftp commands
o Corrected bugs
1.1 THE SUBNET UTILITY
TCP and IP are protocols defined by the Defense Advanced Research Projects
Agency (DARPA) to permit communication between networks using different
protocols and transmission media. DARPA refers to this overall network of
different networks as an Internet. One of the most well-known among DARPA
Internets is the ARPANET, a communications network that spans the country.
Version 3.0 1-1 TCP/IP
According to the traditional design of the DARPA protocols, any
communications network can communicate with a DARPA Internet as long as they
supply a unique Internet address for each host within the network. (The DARPA
Internet manages these addresses in network-wide routing tables.) This way,
any host within a network could access the DARPA network.
Since then, many individual network administrators have created their own
internets to address several communication needs. They use internets to:
o Separate sprawling communications networks into several manageable
networks
o Connect geographically-separate Local Area Networks (LANs) with a
high-speed point-to-point link
o Combine different types of LANs such as ETHERNET LANs and DOMAIN
rings
o Ease network congestion by putting heavily-trafficked hosts on
separate cables
The traditional DARPA Internet model does not support the concept of
individual network users creating their own internets. That is, even though
a communications network may belong to a larger internet, each network
within an internet needs a unique network number to remain on a DARPA
Internet. So, to communicate between two internets on a DARPA Internet, the
hosts must know the network topology of the other.
Consider, for example, two hosts on the ARPANET -- one at the University of
Southern California (USC) and the other at Massachusetts Institute of
Technology (MIT). To send a message from the USC host to the MIT host, the
USC host must specify the appropriate network within the internet at MIT.
Also, any changes to a network within an internet affect DARPA Internet
routing tables. So, any time network administrators change networks within
their internets, they must update the DARPA routing tables.
The subnet utility provides more flexibility in the network structure by
allowing network administrators to subdivide their network without affecting
the DARPA Internet. This way, administrators can keep their network
activities separate from the entire DARPA Internet.
Referring to our example, the USC host can now send a message to host at MIT
by specifying a network and host number. The network number represents the
entire MIT internet. When the message reaches the MIT gateway, the gateway
checks whether subnets are implemented, and if so, relays the message to the
appropriate network within the MIT internet.
TCP/IP 1-2 Version 3.0
To implement a subnet utility, you don't have to use a different DARPA
Internet addressing mechanism. You simply specify a new interpretation of the
current Internet address by supplying a subnet_mask.
Currently, you supply a 32-bit Internet address that identifies each host
(workstation) on your network. The subnet utility allows you to specify a
subnet address for each LAN within your network.
For each host on your network, you specify a unique 32-bit Internet address.
DARPA defines three types of Internet address: A, B, or C. You can
distinguish which type of address is in use by the size of each field. That
is,
o Type A addresses have a 7-bit network number, a 24-bit host number
and the value of the most significant (leftmost) bit is
0.
o Type B addresses have a 14-bit network number, a 16-bit host number
and the value of the two most significant (leftmost) bits are 10.
o Type C addresses have a 21-bit network number, an 8-bit host number
and the value of the three most significant (leftmost) bits are
110.
Figure 1 shows how a 32-bit Internet address is divided into network and
host numbers. The M refers to the most significant bit
field.
01 8 16 24 32
+-+------+--------------------------+
Type A |M| net | host |
+-+------+--------------------------+
2
+--+--------------+-----------------+
Type B |M | net | host |
+--+--------------+-----------------+
3
+---+----------------------+--------+
Type C | M | net | host |
+---+----------------------+--------+
Figure 1. Type A, B, and C Internet Addresses
When using the subnet utility, you can further subdivide the Internet
address into network, subnet and host field. Note that the size of the
network address remains the same, the host field is divided into subnet and
host fields.
Version 3.0 1-3 TCP/IP
Figure 2 shows some possible ways you can subdivide an Internet address into
network, subnet and host numbers.
0 8 16 24 32
+--------+--------+-----------------+
Type A | net | subnet | host |
+--------+--------+-----------------+
OR
+--------+-----------------+--------+
| net | subnet | host |
+--------+-----------------+--------+
+-----------------+--------+--------+
Type B | net | subnet | host |
+-----------------+--------+--------+
OR
+-----------------+------------+----+
| net | subnet |host|
+-----------------+------------+----+
+--------------------------+---+----+
Type C | net |sub|host|
+--------------------------+---+----+
Figure 2. Internet Addresses with Subnet Fields
When determining your Internet addresses, you don't need to remember the size
of each field. Instead, you can simply be sure to specify a number within
the given range. The range of an the Internet address is represented in
decimal number values. That is, the 4-byte Internet address is represented
by four decimal numbers within the range of 0 and 255.
For example, Type C addresses have a one-byte host address, so you can
choose any number between 1 and 254. (DARPA Internet reserves 0 and 255.)
The network address is 3-bytes long and you can choose any number between
192.0.1 through 223.255.254. This number starts after 192 because the first
three bits (0 through 192 in decimal) are reserved to signify the Type C
address.
TCP/IP 1-4 Version 3.0
Table 1 summarizes the range of values you can specify for Type A, B and C
addresses.
TABLE 1. Range of Network and Host Values
for Type A, B, and C Addresses
-------------------------------------------------------------------
Type Size in Bytes Range of Values
Network Host Network Host
-------------------------------------------------------------------
A 1 3 1 - 126 0.0.1 - 255.255.254
B 2 2 128.1 - 191.254 0.1 - 255.254
C 3 1 192.0.1 - 223.255.254 1 - 254
-------------------------------------------------------------------
To create a subnet, you subdivide the host portion of your Internet address.
Table 2 lists the range of subnet and host values for each type. Note that
since Type C host numbers are only 8-bits long, you're limited to 15 subnets
and 14 hosts. For this reason, most users implement subnets with Type A or
B addresses.
TABLE 2. Range of Subnet and Host Values
for Type A, B, and C Addresses
-------------------------------------------------------------------
Type Size in Bits Range of Values
Subnet Host Subnet Host
-------------------------------------------------------------------
A 16 8 0.1-255.255 1 - 254
A 8 16 1-255 0.1- 255.254
B 8 8 1 - 255 1 - 254
C 4 4 1 - 15 1 - 14
-------------------------------------------------------------------
As stated previously, when implementing subnets you are merely changing the
interpretation of your Internet address by supplying a bit mask or subnet
mask. The mask identifies which bits of the Internet address correspond to a
subnet number, and which bits correspond to the host number.
To supply the mask, you edit your Internet networks file
(/sys/node_data/[.node_id]/networks). You must supply the following:
o A semicolon to separate the mask information from the physical
interface information
o The word, mask
o Your Internet address with network and subnet fields denoted by one's
(255), and host field denoted by zero's (0)
Version 3.0 1-5 TCP/IP
For example, the following is a network entry without a subnet mask.
NOTE: Previously, TCP/IP releases referred to the ETHERNET interface as
il. This has been changed to eth. Version 3.0 will accept either eth or il,
however future revisions will accept only eth.
This example indicates that you have a Type A internet address, on an
ETHERNET (eth) interface. We know this is a Type A address because the first
number is within the range of 1 and 126.
10.9.9.7. on eth0
The following is a network entry with a subnet mask. Given that this is a
Type A address, we know that the first field is the network number. The next
field is the subnet number because it is all one's, and the host number
corresponds to the last two bytes, as indicated by zeros.
10.9.9.7 on eth0; mask 255.255.0.0
The following is a two-byte subnet mask for a Type A address:
10.9.9.7 on eth0; mask 255.255.255.0
The following is a one-byte subnet mask for a Type B address where the first
two bytes indicate the network number, the third byte is the subnet number,
and the fourth byte is the host number.
129.9.9.9 on eth0; mask 255.255.255.0
The following is a Type C address with a 4-bit subnet and 4-bit host field.
195.9.9.7 on eth0; mask 255.255.255.240
For more information on specifying Internet addreses and editing the networks
file, see the Configuring_and_Managing_TCP/IP manual (008543).
TCP/IP 1-6 Version 3.0
1.2 CHANGES TO tcpstat
The TCP/IP tcpstat command with the -i option, which reports information
about physical interfaces changed with this release. Ihe command now
displays an additional field, MASK, to report subnet masks. It also changed
the way it reports status in the STAT field. We describe these changes in the
sections below.
1.2.1 tcpstat TO REPORT SUBNETS
You can determine whether your network has a subnet utility by typing the
tcpstat -i command. This command now displays a field containing the
subnets mask for each physical network interface. This network mask
corresponds to the information provided in the networks file.
Figure 3 is a sample report. (For printing purposes, we split this report in
two.) It indicates that the first item is a DOMAIN ring interface (dr0) with
a Type C Internet address and no subnets. The second item is an ETHERNET
interface (eth0) with a Type A Internet address and no subnets. The third
item is an ETHERNET interface (eth0) with a Type A Internet address, a
one-byte subnet field, and a two-byte host field.
$ tcpstat -i
UNIT ADDRESS STAT IPKTS OPKTS RSTS FLSH
dr0 192.9.10.31 AI 1703 1460 7 0
eth0 127.0.0.1 AI 0 0 0 0
eth0 10.9.9.7 AI 0 0 0 0
OERR IERR COLL MASK
7 0 0 255.255.255.0
0 0 0 255.0.0.0
0 0 0 255.255.0.0
$
Figure 3. Generating a Status Report with tcpstat -i
Note that you cannot get this information about subnet masks with DOMAIN/IX
BSD4.2 TCP/IP. The corresponding command, netstat, does not report information
on the subnet utility.
Version 3.0 1-7 TCP/IP
1.2.2 tcpstat CHANGE TO STAT FIELD
The tcpstat -i status report, which reports the status flags for each
physical interface, has been improved with this release. It
reports more status codes, with mnemonics rather than bits. Table 3
lists the status codes and what they mean. Note that AI indicates that
the interface is healthy. E indicates the interface is working, though
some errors have ocurred.
TABLE 3. tcpstat Status Codes
-------------------------------------------------------------------------------
Status Meaning Status Meaning
Code Code
-------------------------------------------------------------------------------
A available G global
C initialization completing I initialized
D disabled M subnets in use
E error W waiting for initialization
F flushing
-------------------------------------------------------------------------------
1.3 CHANGES TO RUNNING tcpserver with the DEBUG OPTION
You can run the tcpserver in a window with the debug option to troubleshoot
TCP/IP. Prior to this revision, the debug mode displayed all its activity in
the window. With Version 3.0, you can control what types of debug
information tcpserver displays during the debug session by specifying
values on the command line. This allows you to suppress debug information
about activities that you don't need to see.
To get available debug information, run tcpserver in a window with the
-debug option on the command line as follows:
$ /sys/tcp/tcpserver -debug [hexadecimal value]
The hexadecimal value you specify corresponds to a 16-bit mask. If the bit
is set, the corresponding information will be displayed. The bits are
defined as follows:
TCP/IP 1-8 Version 3.0
Bit Debug Information
0001 General information
0002 IP level information
0004 ARP information
0008 TCP information
0010 Data in TCP packets
0020 UDP information
0200 Broadcasts
1000 TCP Finite State Machine information
2000 Device level information
4000 Additional detail at any level
If you specify the -debug option without any hexadecimal values, you'll get
general information. To get additional information, you can specify bit
values corresponding to other debug information.
For example, to specify TCP (0008) and IP (0002) information, you add the
bits 0002 and 0008 to specify the following command line:
$ /sys/tcp/tcpserver -debug 000a
To specify TCP, IP, and device level (2000) information, you add the bits
0002, 0008, and 2000. So you specify the following command line:
$ /sys/tcp/tcpserver -debug 200a
Controlling which information gets displayed is often quite helpful during
troubleshooting. For example, you might want to get all the available debug
information except for broadcast information. (You might want to suppress
broadcast information when routed or ripserver are running on the local
network because they generate many broadcasts.) To get all the available
debug information execpt broadcast information, supply the following bitmask
on the command line:
$ /sys/tcp/tcpserver -debug f0ff
Version 3.0 1-9 TCP/IP
1.4 CHANGE TO TCP/IP INSTALLATION: TCP/IP NO LONGER CONTAINS DRIVER
Prior to this release, DOMAIN TCP/IP provided driver software that enabled
TCP/IP to run over ETHERNET gateways. This version of TCP/IP does not
provide driver software. Instead, you must first install the appropriate
driver before installing or updating TCP/IP. You can refer to the following
release documents for installing these drivers:
o EtherController-AT Release Document (009742) for installing an
ETHERNET controller in DN3000 workstations and servers.
o EtherController-MB Release Document (009743) for installing an
ETHERNET controller in workstations and servers that contain the
MULTIBUS peripheral expansion cage.
NOTE: Previously TCP/IP releases referred to the ETHERNET
interface as il. This has been changed to eth. Version 3.0 will
accept either eth or il, however future revisions will accept only
eth.
1.5 LARGER UDP PACKET SIZE
TCP/IP Version 2.1 supported a maximum User Datagram Protocol (UDP) packet
size of 1024 bytes. Version 3.0 supports a UDP size of up to 9132 bytes.
1.6 NEW telnet DEBUG COMMAND
You can now run telnet in debug mode by specifying the -debug option when
you invoke telnet as follows:
$ telnet -debug
When you run telnet in debug mode, telnet generates a packet trace for the
connection.
TCP/IP 1-10 Version 3.0
1.7 ADDITIONAL ftp COMMANDS
The DOMAIN File Transfer Program (ftp), which allows you to transfer files
between your DOMAIN network and a reomote host, has added new commands and
new functions to existing commands. Table 4 lists the additions and changes
to ftp commands. The following sections describe these commands in more
detail.
TABLE 4. Additional ftp Commands
-------------------------------------------------------------------------------
Command Description Command Description
--------------------------------------------------------------------------------
mget/mrecv/mretrieve Copy remote files to mkdir Make directory on
local host remote host
mput/msend/mstore Copy local files to pwd Display remote
remote host working directory
put Same as ftp store rmdir Delete directory on
and send command remote host
recv Same as ftp get, wd/cd List local working
retrieve command directory
mdelete Delete multiple files debug Toggle debugging mode
on remote host
cup Move to remote parent remotehelp Display remote host
directory available commands
ld Display local verbose Toggle verbose mode
working directory
acct Same as ftp account command/cmd Create a Shell process on
command. the remote host.
trace Toggle trace mode to
generate packet traces.
-------------------------------------------------------------------------------
Version 3.0 1-11 TCP/IP
1.7.1 THE mget, mrecv, mretrieve and COMMANDS
----------------------------------------------------------------------
mget/mrecv/mretrieve -- Copy remote files to local host
----------------------------------------------------------------------
FORMAT
mget [OR mrecv OR mretrieve] remotefiles -option
With the mget command (and its synonyms mrecv and mretrieve), you can use
wildcards when copying remote files to a local host. You can copy the files
using one of the following options on the command line:
Option Action
-q Query for pathname of destination
-s Destination pathname same as source pathname
-w Source file copied to current working directory
ARGUMENTS
remotefiles
Specify the names of the files to be copied to the local host. If you omit
the argument on the command line, the command will prompt you for
it.
EXAMPLE
> mretrieve /src/a* -w
getting list of names from server....
200 Host 192.9.10.96, port 2263
125 List started ok
250 List completed
retrieve "/src/a_dest" as "a_dest"
200 Host 192.9.10.96, port 2366
150 Retrieval of "/src/a_dest" started
226 Transfer of "/src/a_dest" completed
retrieve "/src/a_dst2" as "a_dst2"
200 Host 192.9.10.96, port 2889
150 Retrieval of "/src/a_dst2" started
226 Transfer of "/src/a_dst2" completed
> ld
Directory "/yoyo":
a_dest a_dst2
2 entries
TCP/IP 1-12 Version 3.0
If you don't supply the remote files, the command will prompt for them.
> mretrieve
remote file-group descriptor: /src/a_*
flags <-q/-s/-w>: -w
.
.
.
1.7.2 THE mput, msend, and mstore COMMANDS
----------------------------------------------------------------------
mput/msend/mstore -- Copy local files to a remote host
----------------------------------------------------------------------
FORMAT
mput [OR msend OR mstore] localfiles -option
With the mput command (and its synonyms msend and mstore), you can use
wildcards when copying local files to a remote host. You can copy the files
using one of the following options on the command line:
Option Action
-q Query for pathname of destination
-s Destination pathname same as source pathname
-w Source file copied to current working directory
ARGUMENTS
localfiles
Specifies the names of the files to be copied to the local host. If you omit
the argument on the command line, the command will prompt you for it.
EXAMPLE
> mput /y2/*dst* -s
200 OK
200 Host 192.9.10.96, port 3094
125 Storing "a_dst2"
226 File transfer completed
200 OK
200 Host 192.9.10.96, port 3430
Version 3.0 1-13 TCP/IP
125 Storing "a_dst2_a"
226 File transfer completed >
1.7.3 THE put COMMAND
The ftp put command is the same as the ftp commands store and send. The
command allows you to send a local file to a remote host. For a complete
description, see the section on store and send in the manual, Using_telnet
and_ftp.
1.7.4 THE recv COMMAND
The ftp recv command is the same as the ftp commands get and retrieve. The
command allows you to transfer a remote file to a local host. For a complete
description, see the section on retrieve and get in the manual, Using__telnet
and_ftp.
1.7.5 THE mdelete COMMAND
----------------------------------------------------------------------
mdelete -- Delete multiple files on the remote host
----------------------------------------------------------------------
FORMAT
mdelete remotefiles
Use the mdelete command to delete several files using wildcards. The command
queries you before it deletes a file. Supply the following responses to each
query:
Response Action
y Delete the file
n Save the file
q Stop mdelete, return to ftp '>' prompt
g Delete files without further queries
ARGUMENTS
remotefiles
Specify the files you want to delete using DOMAIN wildcards.
TCP/IP 1-14 Version 3.0
EXAMPLE
> mdelete /yoyo/a*
getting list of names from server....
200 Host 192.9.10.96, port 2519
125 List started ok
250 List completed
Delete file '/yoyo/a_dest' <y/n/g/q>? y
250 "/yoyo/a!dest" deleted
Delete file '/yoyo/a_dst2' <y/n/g/q>? n
>
1.7.6 THE cup COMMAND
----------------------------------------------------------------------
cup -- Move to the parent direcory on the remote host
----------------------------------------------------------------------
FORMAT
cup
The cup command changes the working directory on the remote host to the
parent directory.
1.7.7 THE ld COMMAND
----------------------------------------------------------------------
ld -- Displays the local working directory
----------------------------------------------------------------------
The ld command allows you to display the working directory on the local
directory.
FORMAT
ld
Version 3.0 1-15 TCP/IP
1.7.8 THE mkdir COMMAND
----------------------------------------------------------------------
mkdir -- Make a directory on the remote host
----------------------------------------------------------------------
FORMAT
mkdir pathname
The mkdir command creates a new directory named pathname on the remote host.
ARGUMENTS
pathname
Specify the name of the directory you are creating.
1.7.9 THE pwd COMMAND
----------------------------------------------------------------------
pwd -- Display the current remote working directory
----------------------------------------------------------------------
FORMAT
pwd
The pwd command displays the name of the current working directory on the
remote host.
TCP/IP 1-16 Version 3.0
1.7.10 THE rmdir COMMAND
----------------------------------------------------------------------
rmdir -- Delete a directory on the remote host
----------------------------------------------------------------------
FORMAT
rmdir pathname
The rmdir command deletes the directory named pathname on the remote host.
The directory must be empty; you must have previously removed all the files
with the delete or mdelete command before you can remove the directory.
ARGUMENTS
pathname
Specify the name of the directory you want to delete.
1.7.11 THE wd/cd COMMAND
----------------------------------------------------------------------
wd/cd -- Changes the local working directory
----------------------------------------------------------------------
FORMAT
wd [OR cd] pathname
The wd or cd command changes the working directory on the local host to the
directory specified by pathname. If you don't specify a pathname, wd or cd
displays the local working directory.
Version 3.0 1-17 TCP/IP
1.7.12 THE acct COMMAND
The ftp acct command is the same as the ftp account command. This command
allows you to send your account number to the remote host. For a complete
description, see the section on account in the manual, Using__telnet__and
ftp.
1.7.13 THE command/cmd COMMAND
----------------------------------------------------------------------
command/cmd -- Changes the local working directory
----------------------------------------------------------------------
FORMAT
command [OR cmd] [shellcommand [commandargs]]
The command or cmd command invokes a DOMAIN Shell process and passes
commands to this process rather than the executing process.
If you enter a shellcommand immediately after command, the Shell process
exists after processing the command. If you enter the command without
specifying a shellcommand, the DOMAIN Shell prompt ($) appears, and you can
enter any number of Shell commands. You exit the Shell process by entering a
EOF or CTRL/Z command.
Note that connections to the remote host remain open while running the Shell
process.
ARGUMENTS
shellcommand
Enter the DOMAIN Shell command that you want to run on the local machine.
Default if omitted: DOMAIN Shell prompt appears and the Shell will process
commands up to an EOF or CTRL/Z.
commandargs
Specify any arguments required to run shellcommand.
TCP/IP 1-18 Version 3.0
1.7.14 THE debug COMMAND
----------------------------------------------------------------------
debug -- Toggle debugging mode
----------------------------------------------------------------------
FORMAT
debug
The debug command turns debugging on and off. When debugging is on, ftp
prints each command that it transmits to the remote ftp server, preceded by
the string "-->". By default, debug mode is off.
1.7.15 THE remotehelp COMMAND
----------------------------------------------------------------------
remotehelp -- Display available commands on remote host
----------------------------------------------------------------------
FORMAT
remotehelp [command]
Use this command to see a description of the ftp commands available on the
remote host. If you want information about a specific command, type the
name of the command on the command line. You will receive information about
that command, if it's available on the remote host.
ARGUMENTS
command
Specify the ftp command you want information about. If you omit this,
remotehelp displays a list of all available ftp commands on the remote host.
Version 3.0 1-19 TCP/IP
1.7.16 THE trace COMMAND
----------------------------------------------------------------------
trace -- Toggle trace mode
----------------------------------------------------------------------
FORMAT
trace
The trace command turns tracing on and off. When tracing is on, ftp
generates a packet trace for the connection. By default, trace mode is off.
1.7.17 THE verbose COMMAND
----------------------------------------------------------------------
verbose -- Toggle verbose mode
----------------------------------------------------------------------
FORMAT
verbose
When you run in verbose mode, the remote ftp server displays all its
responses on the local host. It also displays statistics regarding the
efficiency of data transfer. By default, verbose is on.
1.7.18 CHANGE TO THE list and nlst COMMANDS
You can now direct a listing to standard output with the commands list and
nlst by supplying a hyphen ('-') rather than a pathname as a second argument.
list displays information about a remote file (specified in the third
argument), and nlst displays information about a list of remote filenames.
For more information, see the manual, Using_telnet_and_ftp.
TCP/IP 1-20 Version 3.0
CHAPTER 2
SOFTWARE INSTALLATION PROCEDURES
INSTALLATION PROCEDURE
This chapter describes how to install TCP/IP Version 3.0. You can add this
software to a user node (one equipped with monitor and keyboard) or a DOMAIN
server processor (DSP) that is running SR9.5 of the AEGIS or DOMAIN/IX
operating system. If the user node or DSP is not running SR9.5 or a more
recent version, follow the appropriate software update procedures as
described in Installing__DOMAIN__Software (008860) or in the appropriate
release notes.
Prior to this release, DOMAIN TCP/IP provided driver software that enabled
TCP/IP to run over ETHERNET gateways. This TCP/IP does not provide driver
software. Instead, you must first install the appropriate driver before
installing or updating TCP/IP. You can refer to the following release
documents for installing these drivers:
o EtherController-AT Release Document (009742) for installing an
ETHERNET controller in DN3000 workstations and servers.
o EtherController-MB Release Document (009743) for installing an
ETHERNET controller in workstations and servers that contain the
MULTIBUS peripheral expansion cage.
If you have both the DOMAIN/IX and TCP/IP products, you must install
DOMAIN/IX Version 9.5 before you install TCP/IP. See the Release Document for
DOMAIN/IX Version 9.5 for instructions on how to install this software.
NOTE: The user node or DSP must have a minimum of 800 blocks of available
disk space for a successful installation of this software.
NOTE: If you're running TCP/IP as a gateway on a DOMAIN node containing
the EtherBridge product, you must update all the nodes running
TCP/IP on the network to Version 3.0 if you want the nodes to
communicate through the gateway. For details on using EtherBridge
and TCP/IP, see the EtherBridge Version 1.3 Release Notes.
Version 3.0 2-1 TCP/IP
NOTE: TCP/IP Version 3.0 is not compatible with SR8 software. Therefore
nodes running SR8 software cannot use the
gateway.
2.1 Conventions and Terminology
Before you start, make sure you understand these terms and conventions:
Work Node The user node at which you perform the installation
procedure.
Target The directory into which you're installing software. The
target can be a node entry directory (for example, //TARGET)
or any subdirectory (for example, //TARGET/PRODUCT). If the
target is on a user node, then the work node and the user
node can be the same node.
NOTE: When you are installing software to update a
diskless node, the target is the node entry
directory of the partner node.
Secure network A network that uses a registry of user accounts and access
control lists (ACLs) to control log-in privileges and access
to files and directories. Note that an open network does not
use a registry or ACLs.
Source area An on-line master area of DOMAIN software. An administrator
installs software from the distribution media into the source
area and users install software from the source area over the
network. The source area can be a node's entry directory or
any subdirectory.
Source media The media (floppy disks, magnetic tape, cartridge tape, or
another node in the network) that contains the software.
< > Angle brackets ( < > ) enclose the name of a key on the
keyboard.
2.2 Planning the Installation
There is one installation procedure. You can use the procedure in one of
three modes: ADMINISTRATIVE mode, USER mode, or SPECIAL-CASE mode.
ADMINISTRATIVE mode creates a source area by copying the INSTALL program and
the new software from the distribution media to the target. Use the
administrative mode to provide a source area for this release. See Section
2.3 for more information about administrative-mode installations.
TCP/IP 2-2 Version 3.0
USER mode involves copying your new software from a source area onto another
node in the network; it's the simplest and most commonly used mode. You can
install in user mode only AFTER an administrative-mode installation has
initialized the source area with the INIT_SOURCE program.
Two default conditions apply to a user-mode installation. The defaults are:
o The INSTALL program automatically copies the new software over the
network from the initialized source area, instead of asking you to
specify the source area.
o The INSTALL program uses the SID "user.sys_admin" during the
installation, rather than your own login SID.
To install in user mode, get the source area's pathname from your system
administrator, then go on to Section 2.4.2.
SPECIAL-CASE mode involves special cases in which you need to override the
user-mode defaults. The special cases are:
o You want to install software from an initialized source area on the
network, but your own login SID gives you more rights to a target's
protected directories than the default SID "user.sys_admin"
o You want to install software from a source other than an initialized
source area (for example, source media)
o You want to install additional software in a source area that was
initialized during a previous administrative-mode installation
The installation procedure enters special-case mode when you invoke the
INSTALL program with its -my_sid option. Specifying this option overrides
the user-mode defaults, which means that the INSTALL program (1) uses your
own login SID instead of "user.sys_admin" and (2) prompts for source media
rather than automatically copying the software over the network from an
initialized source area. In all other respects, special-case mode behaves
like one of the other two modes of installation (your choice of source
determines which one).
If you want to install software from an initialized source area on the
network while using your own login SID, follow the directions for user-mode
installations in Section 2.4.2. These directions include provisions for
installing software in special-case mode.
If you want to install software from a source other than an initialized
source area, or you want to install additional software in a previously
initialized source area, follow the directions for an administrative-mode
installation in Section 2.4.1. These directions also include provisions for
installing software in special-case mode.
Version 3.0 2-3 TCP/IP
2.3 More Information About Administrative-mode Installations
The target of an administrative-mode installation generally serves as the
source for subsequent user-mode installations (the administrative-mode target
pathname is therefore the same as the user-mode source pathname). User-mode
installations use both the INSTALL program and the software stored in the
source area.
ADMINISTRATIVE MODE USER MODE
SOURCE --------------------> TARGET (user node or DSP)
______
(media) SOURCE -------------------> TARGET
Your choice of target for an administrative-mode installation depends on
whether you want the target node to RUN the software as well as act as a
SOURCE for the software. If you also want the node to run the software, make
the target the node's entry directory (for example, //node). If you just
want the node to contain the software, you should make the target a
subdirectory (for example, //node/product/source_area). In either case,
users should then use the target of your administrative-mode installation as
their source area.
You can install different optional software products into the same source
area or into separate source areas. Whichever route you take, you can then
selectively install optional products on user nodes or DSPs from the source
area(s).
If you have a secure network, you must have system administrator rights to
install in administrative mode. Also, during the procedure you must
initialize the source area by running the INIT_SOURCE program. This program
marks the installation program in the source area with special privileges for
subsequent user-mode installations, such as use of the SID "user.sys_admin"
during installation. The INSTALL program can then install software in
protected system directories, even though the user running the program does
not have rights to modify these directories. In open networks, you create a
source area but you don't run INIT_SOURCE, since all users have rights to
modify their system directories.
2.4 Installation Procedures
The following sections describe the administrative mode and the user mode of
installation. To install software in special-case mode, consult Section 3.0
to determine which set of instructions you should follow.
TCP/IP 2-4 Version 3.0
2.4.1 Administrative Mode
NOTE: You can enter "q" or "quit" at any prompt in the INSTALL program to
abort the installation and return to the Shell.
1. If you intend to create a source area for future installations, log
on to a work node using a system administrator account (for example,
my_name.sys_admin.%.%). Otherwise, log on using your own account
(for example, my_name.%.%.%).
2. Set your working directory to the installation target. This target
will become the source area for user installations. It can be a node
entry directory (like //node) or it can be any subdirectory created
prior to the installation (like //node/product/source_area). For
example:
$ wd //node
3. Insert the source media into the drive and enter the RBAK command
shown below. If you are using a tape cartridge, use the CT option
shown in the example. If you are using a magnetic tape, use the M0
(Mzero) option. If you are using a floppy disk, use the F0 (Fzero)
option.
NOTE: TCP/IP comes on multiple floppy disks. Insert the floppy
disk with the numeral 1 at the end of its label (for
example, "FLP8_TCP_3.0_1"). You will perform this step
only with the first floppy disk.
All of the RBAK commands shown below create an INSTALL directory on
the target and write the installation software to the directory.
When entering the RBAK command, use lower-case characters to ensure
visibility of the install directory in case-sensitive environments.
Note that you can leave the source media in the drive for use in a
later step; if you remove the source media after executing the RBAK
command, the INSTALL program will later prompt you to re-insert the
media.
$ rbak -dev ct -f 1 install -as install -l -ms -force -sacl -du
$ rbak -dev m0 -f 1 install -as install -l -ms -force -sacl -du
$ rbak -dev f0 -f 1 install -as install -l -ms -force -sacl -du
Version 3.0 2-5 TCP/IP
4. Set your working directory to the INSTALL directory on the target.
For example:
$ wd //node/install
5. Execute the INSTALL program and follow the prompts. If you are
installing software in special-case mode, use the -my_sid option.
For ADMINISTRATIVE MODE, type:
$ install
For SPECIAL-CASE MODE, type:
$ install -mysid
6. The install program may prompt you to enter an installation type,
based on what products already exist in the source area. If it
does, answer OPT and proceed. For example:
****************************************
* SOFTWARE INSTALLATION -- Version n.n *
****************************************
Software installation TYPES are:
STD -- Install standard software
RESTART -- Restart the software installation
ACL -- Set ACLs for existing software
CLEANUP -- Run the Cleanup Procedure for ADD MODE installations
OPT -- Install optional software (e.g., Pascal, FORTRAN)
Please enter installation TYPE: OPT
TCP/IP 2-6 Version 3.0
7. When the program displays the names of one or more optional
products, enter the name of the optional product that you want to
install. For example, to install TCP/IP, type "TCP", as shown in the
sample menu below.
Name Description Disk Blocks Needed
(Adding New Software)
TCP TCP/IP 800
OTHER If the optional product that you would like to install
is not listed above, choose OTHER.
*Note: When you choose OTHER, you are asked a few questions
then shown a display of Apollo's optional products. Check
with your system administrator to determine which products
your site has purchased and in which directory these
products have been installed.
Enter the name of a single product you would like to install: TCP
8. When prompted for the name of the target, enter the appropriate
pathname (that is, the node entry directory that you specified in
Step 2). Note that you must install TCP in the node entry
directory. For example:
The TARGET is the node on which you are installing
software. (e.g., '//my_node' or '//my_node/subdirectory')
Enter Target: //node
9. Indicate whether you are installing software on a gateway to another
network. Answer YES if you are installing software on any of the
following:
o On a gateway to another network
o On a bridge to a DOMAIN network
o In a source area that will contain all TCP/IP software
Is the installation to //TARGET a GATEWAY install ?
Enter YES or Y or NO or N :
Version 3.0 2-7 TCP/IP
10. Indicate whether you are installing software for a diskless node.
Is the installation to //TARGET a DISKLESS install ?
Enter YES or Y or NO or N :
If you specify YES, the procedure asks you for the ID of the
diskless node. Enter the hexadecimal node ID.
Please enter the NODE ID of the TARGET DISKLESS NODE that the tcp
software will be installed for. (e.g., 260d):
Enter hex node id:
11. Indicate whether this is a TCP/IP administrative node that will
maintain a copy of the host mapping files. If so, answer YES. Also
answer YES if you are creating a source area that will contain all
TCP/IP software.
Is the installation to //TARGET an TCP/IP ADMINISTRATOR install ?
Enter YES or Y or NO or N :
12. The following question appears if you did NOT specify a GATEWAY
installation in step 9. Indicate whether this node is a client,
that is a TCP/IP host.
Is the installation to //TARGET a CLIENT install ?
Enter YES or Y or NO or N :
13. If you did NOT specify an administrative installation in step 11,
the INSTALL program asks for the TCP/IP administrative node's name.
Please enter the name of the TCP/IP ADMINISTRATOR node on which
the sys/tcp/hostmap DIRECTORY resides (e.g. //SERVER)
Enter node name or type 'quit':
14. The INSTALL program prompts for the source media. Enter your
choice.
Source MEDIA is one of:
CTAPE -- Cartridge Tape
MTAPE -- Magnetic Tape
FLOPPY -- 8" or 5 1/4" Floppies
NET -- An area on the network with valid Software
Enter Source Media:
TCP/IP 2-8 Version 3.0
15. The INSTALL program may ask you to insert the media into the drive.
Insert the media and press <RETURN>.
16. The INSTALL program installs the software, listing each file it
copies from the source media. Since the software resides on
multiple floppy disks, the program prompts you to mount (that is,
insert) the next disk and to press <RETURN> to continue.
When the INSTALL program finishes installing the software, it
displays the following menu:
Options:
RERUN -- There were errors in the transcript
pad and you wish to rerun the installation.
FINISH -- The installation ran to completion error free.
There is no additional optional software you
wish to install.
CONTINUE -- Install additional optional software.
If you encountered any errors during the installation, correct the
problem(s) and select RERUN. To locate error messages issued during
installation, search backwards for the characters @? (an at sign
followed by a question mark) in the installation's transcript pad.
If there were no errors, choose CONTINUE or FINISH. Selecting
CONTINUE brings you back to the beginning of the INSTALL program;
selecting FINISH terminates the program. If you were installing
software from magnetic tape, cartridge tape, or floppy disks, you
can now remove the media from the drive.
17. If you have a secure network and you want the target of your
installation to be used as a source area for future installations,
run the INIT_SOURCE program (also run the program if you are adding
software to a previously initialized source area). You must be
logged in as a system administrator to perform this step.
Invoke INIT_SOURCE at the Shell prompt. When prompted, enter the
pathname of the new source area (which is currently the target of
your administrative-mode installation). Here is an example:
$ initsource
Please enter the name of the SOURCE AREA
for your network (e.g., '//NODE/SOURCE_AREA'):
//node
The source area for your network
Version 3.0 2-9 TCP/IP
has been set to: //node
18. Perform this step only after completing an error-free installation
and selecting FINISH.
a. Use the Display Manager SHUT command to shut down the target
node.
<CMD> SHUT <RETURN>
b. After the SUCCESSFUL SHUTDOWN message and the > prompt
appear, reboot the node by typing the following at the
prompt:
> RE <RETURN>
> <RETURN>
MD REV xx/xx/xx
> EX AEGIS <RETURN>
This is the end of the administrative-mode installation procedure. Refer to
Configuring_and__Managing__TCP/IP for information about configuring your
system.
2.4.2 User Mode
NOTE: You can enter "q" or "quit" at any prompt in the INSTALL program to
abort the installation and return to the Shell.
Follow this procedure if you are installing TCP/IP software on a node
from the network.
1. Log on to a work node using your own account (for example,
my_name.%.%.%).
2. Set your working directory to the INSTALL directory in the source
area (if necessary, ask your system administrator for the pathname).
For example:
$ wd //node/install
TCP/IP 2-10 Version 3.0
3. Execute the INSTALL program and follow the prompts. If you are
installing software in special-case mode, use the -my_sid option.
For USER MODE, type:
$ install
For SPECIAL-CASE MODE, type:
$ install -mysid
4. The program may prompt you to enter an installation type, based on
what products already exist in the source area. If it does, answer
OPT and proceed. For example:
****************************************
* SOFTWARE INSTALLATION -- Version n.n *
****************************************
Software installation TYPES are:
STD -- Install standard software
RESTART -- Restart the software installation
ACL -- Set ACLs for existing software
CLEANUP -- Run the Cleanup Procedure for ADD MODE installations
OPT -- Install optional software (e.g., Pascal, FORTRAN)
Please enter installation TYPE: OPT
5. When the program displays the names of one or more optional
products, enter the name of the optional product that you want to
install. For example, to install TCP/IP you would type "TCP", as
shown in the sample menu below.
Name Description Disk Blocks Needed
(Adding New Software)
TCP TCP/IP 800
OTHER If the optional product that you would like to install
is not listed above, choose OTHER.
*Note: When you choose OTHER, you are asked a few questions
then shown a display of Apollo's optional products. Check
with your system administrator to determine which products
your site has purchased and in which directory these
products have been installed.
Enter the name of a single product you would like to install: TCP
Version 3.0 2-11 TCP/IP
6. When prompted for the name of the target, enter the appropriate
pathname (that is, a node entry directory or subdirectory). For
example:
The TARGET is the node or subdirectory on which you are installing
software. (e.g., '//my_node' or '//my_node/subdirectory')
Enter Target: //node
7. Indicate whether you are installing software on a gateway to another
network. Answer yes if you installing software on a gateway or on a
bridge to a DOMAIN network.
Is the installation to //TARGET a GATEWAY install ?
Enter YES or Y or NO or N :
8. Indicate whether you are installing software for a diskless node.
Is the installation to //TARGET a DISKLESS install ?
Enter YES or Y or NO or N :
If you specify yes, the procedure asks you for the ID of the
diskless node. Enter the hexadecimal node ID.
Please enter the NODE ID of the TARGET DISKLESS NODE that the tcp
software will be installed for. (e.g., 260d):
Enter hex node id:
9. Indicate whether this is an administrative node that will maintain a
copy of the host mapping files.
Is the installation to //TARGET an TCP/IP ADMINISTRATOR install ?
Enter YES or Y or NO or N :
10. The following question appears if you did NOT specify a GATEWAY
installation. Indicate whether this node is a client, that is a
TCP/IP host.
Is the installation to //TARGET a CLIENT install ?
Enter YES or Y or NO or N :
11. If you did NOT indicate that this was an administrative node, the
INSTALL program asks you for the TCP/IP administrative node's name.
Please enter the name of the TCP/IP ADMINISTRATOR node on which
the sys/tcp/hostmap DIRECTORY resides (e.g. //SERVER)
Enter node name or type 'quit':
TCP/IP 2-12 Version 3.0
12. The INSTALL program may prompt for the source area. If so, enter
the pathname (if you don't know it, ask your system administrator).
For example:
The SOURCE AREA is the node or subdirectory from which you are
copying software. (e.g., '//node' or '//node/subdirectory')
Enter Source Area: //node
13. The INSTALL program installs the software, listing the name of each
file it copies from the source area. Upon completion, the INSTALL
program displays the following menu:
Options:
RERUN -- There were errors in the transcript
pad and you wish to rerun the installation.
FINISH -- The installation ran to completion error free.
There is no additional optional software you
wish to install.
CONTINUE -- Install additional optional software.
If you encountered any errors during the installation, correct the
problem(s) and select RERUN (if necessary, consult your system
administrator for assistance). To locate error messages issued
during installation, search backwards for the characters @? (an at
sign followed by a question mark) in the installation's transcript
pad.
If there were no errors, choose CONTINUE or FINISH. Selecting
CONTINUE brings you back to the beginning of the INSTALL program;
selecting FINISH terminates the program.
14. Perform this step only after completing an error-free installation
and selecting FINISH.
a. Use the Display Manager SHUT command to shut down the target
node.
<CMD> SHUT <RETURN>
Version 3.0 2-13 TCP/IP
b. After the SUCCESSFUL SHUTDOWN message and the > prompt
appear, reboot the node by typing the following at the
prompt:
> RE <RETURN>
> <RETURN>
MD REV xx/xx/xx
> EX AEGIS <RETURN>
This is the end of the user-mode installation procedure. Refer to the
Configuring__and_Managing_TCP/IP manual for information about configuring your
system.
TCP/IP 2-14 Version 3.0
2.5 TCP/IP Files
The following sections list the files and links that are installed during a
Client, Server, and Gateway installation.
2.5.1 Client Files
The following files are installed on a client node:
com/ftp
com/host
com/tcpstat
com/telnet
doc/tcp.release_notes
sys/tcp/ftp_server
sys/tcp/makegate
sys/tcp/networks_template
sys/tcp/setroute
sys/tcp/tcpinit
sys/tcp/tcpreset
sys/tcp/tcp_server
sys/tcp/telnet_server
sys/tcp/thishost_template
systest/ssr_util/dtcb
systest/ssr_util/mbd
systest/ssr_util/sodebug
systest/ssr_util/trpt
Version 3.0 2-15 TCP/IP
2.5.2 Gateway Files
The following files are installed on a gateway node in_addition_to all client
node files:
sys/tcp/maphost
sys/tcp/rip_server
2.5.3 Administrative Files
The following files are installed on the administratve node:
sys/tcp/hostmap/hashnic
sys/tcp/hostmap/hosts.txt_template
sys/tcp/hostmap/local.txt_template
sys/tcp/hostmap/makehdb
sys/tcp/hostmap/makehost.sh
sys/tcp/hostmap/ndb_format
sys/tcp/hostmap/sortnic
2.5.4 Links
The installation procedure creates the following links on client and gateway
nodes. //ADMIN indicates the administrative node that you specify during the
installation procedure. TARGET is the target area where the software is
being installed.
FROM TO
TARGET/com/net TARGET/com/hosts
TARGET/sys/tcp/gateways //ADMIN/sys/tcp/gateways
TARGET/sys/tcp/hostmap //ADMIN/sys/tcp/hostmap
TARGET/sys/tcp/hosts.hst //ADMIN/sys/tcp/hosts.hst
TARGET/sys/tcp/networks `node_data/networks
TARGET/sys/tcp/thishost `nod_data/thishost
TCP/IP 2-16 Version 3.0
CHAPTER 3
CHANGES IN DOCUMENTATION
The TCP/IP document set consists of the following manuals:
o Configuring_and_Managing_TCP/IP (008543)
o Using_telnet_and_ftp (008667)
The new features in TCP/IP Version 3.0 have been added to the revised
manual, Configuring__and__Managaing__TCP/IP (008543). This manual was
previously called Managing_TCP/IP-Based_Communications_Products.
Configuring__and__Managing__TCP/IP describes how to configure, manage, and
troubleshoot DOMAIN and DOMAIN/IX BSD4.2 TCP/IP.
Using_telnet_and_ftp describes how to use two common TCP/IP utilities: the
TELNET remote terminal emulator and the FTP file transfer program. This book
describes both the DOMAIN and DOMAIN/IX versions of these utilities.
In addition, the manual, System__Administration___for___DOMAIN/IX___BSD4.2
(009355), contains a subset of information contained in the Configuring_and
Managing_TCP/IP that is relevant to BSD4.2 TCP/IP.
Version 3.0 3-1 TCP/IP
CHAPTER 4
BUGS AND BUG FIXES
4.1 BUGS IN RELEASE 3.0
This section documents known bugs in the TCP/IP documentation, software
installation procedures, and software.
4.1.1 BUGS IN DOCUMENTATION
TCP/IP Version 3.0 comes with a revised manual of Configuring_and_Managing
TCP/IP. Currently, there are no known bugs reported for the documentation.
4.1.2 BUGS IN TCP/IP SOFTWARE
The following bugs currently exist in the TCP/IP software:
o If you send two telnet ip (^ip) commands in a row, with no
intervening input, the second one is received by the foreign host as
the character 't'.
o The telnet ^S sequence and ^Q sequence do not work well because
DOMAIN TCP/IP allows the remote system to transmit up to 8K bytes of
data at a time for performance reasons.
TCP/IP 4-1 Release 3.0
4.2 RESTRICTIONS IN TCP/IP SOFTWARE
The TCP/IP ftp and telnet commands are case-sensitive. You must use
lowercase letters for these commands.
TCP/IP transmits broadcasts using a host address 0. It does, however,
recognize broadcasts from other hosts using broadcast addresses of 0 or -1.
Prior to this release, DOMAIN TCP/IP provided driver software that enabled
TCP/IP to run over ETHERNET gateways. This TCP/IP does not provide driver
software. Instead, you must first install the appropriate driver before
installing or updating TCP/IP. You can refer to the following release
documents for installing these drivers:
o EtherController-AT Release Document (009742) for installing an
ETHERNET controller in DN3000 workstations and servers.
o EtherController-MB Release Document (009743) for installing an
ETHERNET controller in workstations and servers that contain the
MULTIBUS peripheral expansion cage.
Release 3.0 4-2 TCP/IP
4.3 BUG FIXES SINCE RELEASE 2.1
The following TCP/IP bugs have been corrected since TCP/IP Version 2.1:
o Version 3.0 supports Trailer Encapsulations as defined by Request for
Comment (RFC) 893, so you can communicate with TCP/IP implementations
that support trailers. This version corrects a problem with trailers
that occurred in Version 2.1.
o Prior to this release, the telnetserver and ftpserver would try to
initialize before the tcpserver started running and so the servers
failed to start. Version 3.0 corrects this problem. The
telnetserver and ftpserver now wait until tcpserver is running
before they start initializing.
o The telnetserver now negotiates echo options correctly.
o Version 3.0 fixes a problem with the telnetserver that caused it to
hang if you tried to log in to a DOMAIN TCP/IP node, but exited
before you completed the log in procedure.
o Prior to this release, if you passed a bad data buffer address to a
get or put operation, the tcpserver would hang and become
unuseable. Version 3.0 corrects this problem.
o telnet now exits (correctly) when the other side resets the
connection.
o If you logged in from a remote host to a DOMAIN node running DOMAIN
telnet and then tried to run in a C shell, the C shell would hang.
Version 3.0 corrects this.
o If you logged in to a remote system using telnet, and the other side
sent null characters, the null characters would show up as triangles
on our side. Version 3.0 now ignores null characters, so you won't
see any triangles.
o Version 3.0 can now handle TCP/IP windows larger than 32K bytes.
o The ripserver would sometimes stop working properly after it had
been running for a while. That is, it would be running, but wouldn't
do anything. Version 3.0 corrects this.
o Conditional put operations of 4K bytes or more wouldn't work at
times. Version 3.0 corrects this.
o For BSD4.2 TCP/IP users, you can now specify a backlog of 0 in a call
to listen(). This is equivalent to setting a backlog of 1.
TCP/IP 4-3 Release 3.0