📄 rfc46.txt
字号:
Network Working Group Edwin E. Meyer, Jr.
Request for Comments: 46 Massachusetts Institute of Technology
17 April 1970
ARPA Network Protocol Notes
The attached document contains comments and suggestions of the
Network Working Group at Project MAC. It is based upon the protocol
outlined in NWG/RFC 33, 36, and later documents.
This proposal is intended as a contribution to the dialog leading to
a protocol specification to be accepted by the entire Network Working
Group.
We solicit your comments.
I - INTRODUCTION
In this document the Network Working Group at MIT Project MAC suggest
modifications and extensions to the protocol specified by Carr,
Crocker, and Cerf in a preprint of their 1970 SJCC paper and extended
by Crocker in NWG/RFC 36. This document broadly outlines our
proposal but does not attempt to be a complete specification. It is
intended to be an indication of the type and extent of the protocol
we think should be initially implemented.
We agree with the basic concept of simplex communication between
sockets having unique identifiers. We propose the implementation of
a slightly modified subset of the network commands specified in
NWG/RFC36 plus the ERR command as specified by Harslem and Heafner in
NWG/RFC 40.
Given the basic objective of getting all ARPA contractors onto the
network and talking to each other at the earliest possible date, we
think that it is important to implement an initial protocol that is
reasonably simple yet extendable while providing for the major
initial uses of the network. It should be a simple protocol so as to
elicit the broadest possible support and to be easily implementable
at all installations with a minimum of added software.
While the protocol will evolve, the fundamentals of a protocol
accepted and implemented by all installations are likely to prove
very resistant to change. Thus it is very important to make the
initial protocol open-ended and flexible. A simple basic protocol is
more likely to succeed in this respect than a complicated one. This
[Page 1]
RFC 46 ARPA Network Protocol Notes April 1970
does not preclude the existence of additional layers of protocol
between several installations so long as the basic protocol remains
supported.
We feel that three facilities must be provided for in the initial
protocol:
1. Multi-path communication between two existing processes which know
how to connect to each other.
2. A standard way for a process to connect to the logger (logging
process at a HOST) at a foreign HOST and request the creation of a
user process. (The login ritual may or may not be standardized.)
3. A standard way for a newly created process to initiate pseudo-
typewriter communication with the foreign process which requested
its creation.
The major differences between the protocol as proposed by Carr,
Crocker, and Cerf and this proposal are the following:
1. The dynamic reconnection strategy specified in Crocker's
NWG/RFC 36 is reserved for future implementation. We feel that
its inclusion would unduly complicate the initial implementation
of the protocol. We outline a strategy for foreign process
creation that does not require dynamic reconnection. Nothing in
this proposal precludes the implementation of dynamic reconnection
at a later date.
2. We propose that an "instance tag" be added to the socket
identifier so as to separate sockets belonging to different
processes of the same user coexisting at one HOST.
3. The following NCP commands have been added:
a. The ERR command specified in NWG/RFC 40 is included.
b. BLK and RSM commands are presented as possible alternatives to
the "cease on link" IMP command and SPD and RSM commands set
forth in NWG/RFC 36. Because these commands operate on socket
connections rather than link numbers, they do not impede the
implementation of socket connection multiplexing over a single
link number, should that later prove desirable.
c. An INT command that interrupts a process is specified. We feel
that it is highly important to be able to interrupt a process
that may be engaged in unwanted computation or output. To
implement the interrupt as a special format within a normal
[Page 2]
RFC 46 ARPA Network Protocol Notes April 1970
message raises severe difficulties: the connection may be
blocked when the interrupt is needed, and the NCP must scan
each incoming message for an interrupt signal.
d. An ECO echoing command to test communications between NCPs is
included.
4. Sockets are conceptualized as having several states, and these are
related to conditions under which network requests may be queued.
This differs from the unlimited queuing feature, which presents
certain implementation difficulties.
5. The protocol regarding creation of a foreign process and
communication with it is removed to a separate User Control and
Communication (UCC) protocol level and is more fully specified.
II - A HIERARCHY OF PROTOCOLS
It seems convenient and useful to view the network as consisting of a
hierarchy of protocol and implementation levels. In addition to
aiding independent software and hardware development, provisions for
a layered protocol allow additions and substitution of certain levels
in experimental or special purpose systems.
We view the initial network communications system as a hierarchy of
three systems of increasing generality and decreasing privilege
level. These are:
1. IMP Network - The network of IMPs and physical communication lines
is the basic resource which higher level systems convert into more
generalized communication facilities. The IMP network acts as a
"wholesaler" of message transmission facilities to a highly
privileged module within each HOST.
2. Network Control Program - Each HOST contains a module called the
Network Control Program (NCP) which has sole control over
communications between its HOST and the IMP network. It acts as a
"retailer" of the wholesale communications facilities provided by
the IMP network. The network of NCPs can be viewed as a higher
level communications system surrounding the IMP network which
factors raw message transmission capabilities between HOSTs into
communication facilities between ordinary unprivileged processes.
[Page 3]
RFC 46 ARPA Network Protocol Notes April 1970
H O S T A H O S T C
______________________________ ______________________
| | | |
| ____ ____ ____ ____ | | ____ ____ ____ |
| |Proc| |Proc| |Proc| | | | | |Proc| |Proc| | | |
| | A | | B | | C | |UCC | | | | D | | E | |UCC | |
| |____| |____| |____| |____| | | |____| |____| |____| |
| | | | | | | | | | |
- - - - - - |- - - |- - - |- - -|- - -|- - |- - -|- - - |- - - - - -
| | | | | NCP NETWORK | | | |
| | | | | | | | | | |
| _|_____|______|______|_ | | _|_____|______|_ |
| | | | | | | |
| | N C P A | | | | N C P C | |
| |_______________________| | | |________________| |
| || | | || |
|_____________________||_______| |_______||_____________|
|| ||
- - - - - - - - - - - -|| - - - - - - - - - - ||- - - - - - - - - -
|| IMP NETWORK ||
___||___ ____||__
| | | |
| IMP |------------| IMP |
| A | | C |
|________| |________|
| |
| ________ |
| | | |
+------| IMP |-----+
| B |
|________|
FIG 1. Modular View Of Network
3. User Control and Communication Module - The preceding two
communication systems are sufficient to permit communication
between unprivileged processes that already exist. However, one
of the primary initial uses of the network is thought to involve
the creation of a foreign user process through interaction with
the foreign HOST's logger. The User Control and Communication
Module (UCC) implements protocol sufficient for a process to
communicate with a foreign HOST's logger and to make initial
control communication with a created process. Such a process is
to have the same privileges (subject to administrative control) as
a local (to the foreign HOST) user process. The UCC module
communicates through the NCP in a manner similar to an ordinary
process. Except for the ability to close connections to a dead
[Page 4]
RFC 46 ARPA Network Protocol Notes April 1970
process, the UCC module has no special network privileges. The
UCC protocol is only one of several third-level protocols that
could be implemented. For example, a set of batch processing
systems connected through the NCP system might implement a load-
sharing protocol, but not a UCC.
III - NETWORK CONTROL PROGRAM
Each HOST implements a module called the Network Control Program
(NCP) which controls all network communications involving that HOST.
The network of NCPs forms a distributed communication system that
implements communication paths between individual processes. The NCP
protocol issues involve: (i) the definition of these communication
paths, and (ii) a system for coordinating the distributed NCP system
in maintaining these communication paths. These are discussed below.
Sockets
Communication between two processes is made through a simplex
connection between two sockets: a send socket attached to one
process and a receive socket attached to another process. Sockets
have the following characteristics:
Socket Identifier - A socket identifier is used throughout the
network to uniquely identify a socket. It consists of 48 bits,
having the following components:
a. User Number (24 bits) - A socket attached to a process is
identified as belonging to that process by a user number
consisting of 8 bits of "home" HOST code plus 16 bits of user
code assigned by the home HOST. This user number is the same
for all sockets attached to any of his processes in any HOST.
b. Instance Tag (8 bits) - More than one process belonging to a
user may simultaneously exist within a single HOST. The
instance tag identifies the particular process to which a
socket belongs. A user's first process at a HOST to use the
network receives instance tag = 0 by convention.
c. HOST Number (8 bits) - This is the code of the HOST on which
the attached process exists.
d. Socket Code (8 bits) - This code provides for 128 send and 128
receive sockets in each process. The low order bit determines
whether this is a "send" (= 1) or "receive" (= 0) socket.
[Page 5]
RFC 46 ARPA Network Protocol Notes April 1970
States of Sockets - Each socket has an associated state. The NCP may
implement more transitory states of a socket, but the three following
are of conceptual importance.
a. Inactive - there is no currently existing process which has
told the NCP that it wishes to listen to this socket. No other
process can successfully communicate with an inactive socket.
b. Open - Some process has agreed to listen to events concerning
this socket but it is not yet connected.
c. Connected - This socket is currently connected to another
socket.
Socket Event Queue - A queue of events to be disclosed to the owning
process is maintained for each open or connected socket. It consists
of a chronologically ordered list of certain events generated by the
action of one or more foreign processes trying to connect or
disconnect this socket. An entry in the event queue consists of the
event type plus the identifier of the foreign socket concerned. The
following event types are defined:
a. "request" - a foreign socket requests connection. (not queued
if local socket is already connected)
b. "accept" - a foreign socket accepts requested connection.
c. "reject" - a foreign socket rejects requested connection.
d. "close" - a foreign socket disconnects an existing connection.
A "request" event is removed from the queue when it is accepted or
rejected. The other events are removed from the queue as they are
⌨️ 快捷键说明
复制代码
Ctrl + C
搜索代码
Ctrl + F
全屏模式
F11
切换主题
Ctrl + Shift + D
显示快捷键
?
增大字号
Ctrl + =
减小字号
Ctrl + -