rfc891.txt

RFC 相关的技术文档

    corresponding to 1972, bits 5-9 contain the day of the month and

DCN Local-Network Protocols                                        Page 11
D.L. Mills

    bits 10-14 contain the month, starting with one for January.

DATE-VALID
    This is a one-bit state variable used to indicate whether the
    local date and time are synchronized with the master clock.  A
    value of one indicates the local clock is not synchronized with
    the master clock.  This variable is set to one initially and when
    the local time-of-day rolls over past midnight.  It is set to zero
    each time a valid date and time update has been received from the
    master clock. 
    
DELAY
    This is a 16-bit integer temporary variable which represents the
    roundtrip delay in milliseconds to a host.
    
HID
    This is an eight-bit integer temporary variable containing the
    host ID of some host on the local net.
    
    There is a one-to-one correspondence between the Internet
    addresses of local hosts and their HIDs.  The mapping between them
    is selected on the basis of the octet number of the Internet
    address.  For DCN hosts it is the fourth octet, while for hosts
    directly connected to a class-A ARPANET IMP or gateway, it is the
    third octet (logical-host field).  The contents of this octet are
    to be added to ADDRESS-OFFSET to form the HID associated 
    with the address.

HOLD
    This is an eight-bit counter state variable indicating whether
    timestamps are valid or not.  While HOLD is nonzero, timestamps
    should be considered invalid.  When set to some nonzero value, the
    counter decrements to zero at a 1-Hz rate.  Its initial value is
    zero. 
    
HOST-TABLE
    This is a table of NHOSTS entries indexed by host ID (HID).  There
    is one entry for each host in the local net.  Each entry has the
    following format:

    HOST-TABLE.DELAY
        This is a 16-bit field containing the computed roundtrip delay
        in milliseconds to host HID.
        
    HOST-TABLE.OFFSET
        This is a 16-bit field containing the computed signed offset
        in milliseconds which must be added to the local apparent
        clock to agree with the apparent clock of host HID.
        
    HOST-TABLE.PID
        This is an eight-bit field containing the PID of the net-output
        process selected by the routing algorithm to forward packets
        to host HID.
        

DCN Local-Network Protocols                                        Page 12
D.L. Mills


 HOST-TABLE.TTL
     This is an eight-bit field used as a time-to-live indicator.
     It is decremented by the HOST process once each second and
     initialized to a chosen value when a HELLO message is
     received. The table is initialized with the HOST-TABLE.DELAY
     field set to  MAXDELAY for all entries.  The contents of the
     other fields are unimportant. 
  
LOCAL-ADDRESS
    This is a 32-bit bit-string state variable used to contain the 
    local host Internet address.

NET-TABLE
    This is a table of NNETS entries with the following format:

    NET-TABLE.HID
        This is an eight-bit field containing the host ID of the
        pseudo-process to forward packets to the NET-TABLE.NET net.

    NET-TABLE.NET
        This is a 24-bit field containing an Internet class-A (eight
        bits), class-B (16 bits) or class-C (24 bits) net number.
        Note that the actual field width for class-B net numbers is 24
        bits in order to provide a subnet capability, in which the
        high-order eight bits of the 16-bit host address is
        interpreted as the subnet number. 
        
    The table is constructed at configuration time and must include an
    entry for every net that is a potential neighbor.  A neighbor net
    is defined as a net containing a host that can be directly
    connected to a host on the local net.  The entry for such a net is
    initialized with NET-TABLE.NET set to the neighbor net number and
    NET-TABLE.HID set to an arbitrary vitual-host ID not assigned any
    other local-net virtual host. 
    
    The remaining entries in NET-TABLE are initialized at initial-boot
    time with the NET-TABLE.NET fields set to zero and the
    NET-TABLE.HID fields set to a configuration-selected host ID to be
    used to forward packets to all nets other than neighbor nets.  In
    the case where a gateway module is included in the local host
    configuration, the GGP and/or EGP protocols will be used to
    maintain these entries;  while, in the case where no gateway
    module is included, only one such entry is required. 
    
OFFSET
    This is a 16-bit signed integer temporary variable which
    represents the offset in milliseconds to be added to the apparent
    clock time to yield the apparent clock time of the neighbor host. 
    
3.1.2.  Parameters

ADDRESS-OFFSET
    This is an integer which represents the value of the Internet 
    address field corresponding to the first host in HOST-TABLE.

DCN Local-Network Protocols                                        Page 13
D.L. Mills

NHOSTS
    This is an integer which defines the number of entries in HOST-TABLE.

NNETS
    This is an integer which defines the number of entries in MET-TABLE.

3.1.3.  HELLO Packet Fields

PKT.ADDRESS-OFFSET
    This eight-bit is copied from ADDRESS-OFFSET by the sender.

PKT.DATESTAMP
    Bits 0-14 of this 16-bit field are copied from DATE by the sender, 
    while bit 15 is copied from DATE-VALID.

PKT.DATE-VALID
    This one-bit field is bit 15 of PKT.DATESTAMP.

PKT.DESTINATION
    This 32-bit field is part of the IP header.  It is copied from
    HLO.NEIGHBOR-ADDRESS by the sender.

PKT.HOST-TABLE
    This is a table of PKT.NHOSTS entries, each entry of which
    consists of two fields.  The entries are indexed by host ID and
    have the following format: 

    PKT.HOST-TABLE.DELAY
        This 16-bit field is copied from the corresponding HOST-TABLE.DELAY
        field by the sender.

    PKT.HOST-TABLE.OFFSET
        This 16-bit field is copied from the corresponding HOST-TABLE.OFFSET
        field by the sender.

PKT.LENGTH
    This 16-bit field is part of the IP header.  It is set by the sender to
    the number of octets in the packet.

PKT.NHOSTS
    This eight-bit field is copied from NHOST by the sender.

PKT.SOURCE
    This 16-bit field is part of the IP header.  It is copied from
    LOCAL-ADDRESS by the sender.

PKT.TIMESTAMP
    This 32-bit field contains the apparent time the packet was transmitted 
    in milliseconds past midnight UT.


DCN Local-Network Protocols                                        Page 14
D.L. Mills


PKT.TSP
    This 16-bit field contains a variable used in roundtrip delay
    calculations.

3.2 CLOCK Process (CLK)

     The timekeeping system maintains three clocks: (1) the physical
clock, which is determined by a hardware oscillator/counter; (2) the
apparent clock, which maintains the time-of-day used by client
processes and (3) the actual clock, which represents the time-of-day
provided by an outside reference.  The apparent and actual clocks are
maintained as 48-bit quantities with 32 bits of significance available
to client processes.  These clocks run at a rate of 1000 Hz and are
reset at midnight UT.

     The CLOCK process consists of a set of state variables along with
a set of procedures that are called as the result of hardware
interrupts and client requests.  An interval timer is assumed
logically separate from the local clock mechanism, although both could
be derived from the same timing source.

3.2.1.  Local Variables

CLK.CLOCK
    This is a 48-bit fixed-point state variable used to represent the
    apparent time-of-day.  The decimal point is to the right of bit 16
    (numbering from the right at bit 0).  Bit 16 increments at a rate
    equivalent to 1000 Hz independent of the hardware clock.  (In the
    case of programmable-clock hardware the value of CLK.CLOCK must be
    corrected as described below.) 
    
CLK.COUNT
    This is a hardware register that increments at rate R.  It can be
    represented by a simple line clock, which causes interrupts at the
    line-frequency rate, or by a programmable clock, which contains a 16-bit
    register that is programmed to count at a 1000-Hz rate and causes an
    interrupt on overflow.  The register is considered a fixed-point variable
    with decimal point to the right of bit 0.

CLK.DELTA
    This is a 48-bit signed fixed-point state variable used to represent the
    increment to be added to CLK.CLOCK to yield the actual time-of-day.  The
    decimal point is to the right of bit 16.

3.2.3.  Parameters

ADJUST-FRACTION
    This is an integer which defines the shift count used to compute a
    fraction that is used as a multiplier of CLK.DELTA to correct CLK.CLOCK
    once each clock-adjust interval.  A value of seven is suggested.
    

DCN Local-Network Protocols                                        Page 15
D.L. Mills


ADJUST-INTERVAL
    This is an integer which defines the clock-adjust interval in
    milliseconds.  A value of 500 (one-half second) is suggested for
    the line clock and 4000 (four seconds) for the 1000-Hz clock.

CLOCK-TICK
    This is a fixed-point integer which defines the increment in
    milliseconds to be added to CLK.CLOCK as the result of a clock
    tick.  The decimal point is to the right of bit 16.  In the case
    of a line-clock interrupt, the value of CLOCK-TICK should be
    16.66666 (60 Hz) or 20.00000 (50 Hz).  In the case of a 1000-Hz
    programmable-clock overflow, the value should be 65536.00000.
    
HOLD-INTERVAL
    This is an integer which defines the number of seconds that HOLD will
    count down after CLK.CLOCK has been reset.  The resulting interval must be
    at least as long as the maximum HELLO-INTERVAL used by any HELLO process.

3.2.3.  Events and Procedures

INCREMENT-CLOCK Event
    This event is evoked as the result of a tick interrupt, in the case of a
    line clock, or a counter overflow, in the case of the 1000-Hz clock.  It
    causes the logical clock to be incremented by the value of CLOCK-TICK.

    1.  Add the value of CLOCK-TICK to CLK.CLOCK.

ADJUST-CLOCK Event
    This event is evoked once every ADJUST-INTERVAL milliseconds to slew the
    apparent clock time to the actual clock time as set by the SET-CLOCK
    procedure.  This is done by subtracting a fraction of the correction
    factor CLK.DELTA from the value of CLK.DELTA and adding the same fraction
    to CLK.CLOCK.  This continues until either the next SET-CLOCK call or
    CLK.DELTA has been reduced to zero.

    The suggested values for ADJUST-INTERVAL and ADJUST-FRACTION
    represent a maximum slew rate of less than +-2 milliseconds per
    second, in the case of 1000-Hz clock.  The action is to smooth
    noisy clock corrections received from neighboring systems to
    obtain a high-quality local reference, while insuring the apparent
    clock time is always monotonically increasing. 
    
    1.  Shift the 48-bit value of CLK.DELTA arithmetically ADJUST-FRACTION
        bits to the right, discarding bits from the right and saving the
        result in a temporary variable F.  Assuming the decimal point of F to
        be positioned to the right of bit 16 and sign-extending as necessary,
        subtract F from CLK.DELTA and add F to CLK.CLOCK.