📄 rfc1221.txt
字号:
Every HAP message consists of an integral number of 16-bit words (i.e., an even number of octets). The first several words of the message always contain control information and are referred to as the message header. The first word of the message header identifies the type of message which follows. The second word of the message header is a checksum which covers all header information. Any message whose received header checksum does not match the checksum computed on the received header information must be discarded. The format of the rest of the header depends on the specific message type. The formats and use of the individual message types are detailed in the following sections. A common format description is used for this purpose. Words in a message are numbered starting at zero (i.e., zero is the first word of a message header). Bits within a word are numbered from zero (most significant) to fifteen (least significant). The notation used to identify a particular field location is: <WORD#>{-<WORD#>} [ <BIT#>{-<BIT#>} ] <description> where optional elements in {} are used to specify the (inclusive) upper limit of a range. The reader should refer to these field identifiers for precise field size specifications. Fields which are common to several message types are defined in the first section which uses them. Only the name of the field will usually appear in the descriptions in subsequent sections. Link-level protocols used to support HAP can differ in the order in which they transmit the bits constituting HAP messages. The words of the message are transmitted from word 0 to word N.3. Datagram Messages Datagrams are one of the two message types provided by HAP, as described in the previous section. Because network resources are not reserved in advance for datagram traffic, delivery of datagram traffic is subject to greater delivery delays and delay variance than stream traffic, and is subject to flow and congestion controls.Edmond [Page 6]RFC 1221 HAP2 April 1991 Datagram priority determines which packets are delivered or discarded when network resources do not permit handling all of the presented traffic. It is expected that datagram messages will be used to support the majority of computer-to-computer and terminal-to-computer traffic which is bursty in nature. The format of datagram messages and the purpose of each of the header control fields is described in Figure 1. 0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 +--+--+--+--+--+--+--+--+--+--+--+--+--+--+--+--+ 0 | 0|LB|GOPRI| 0 | F| MESSAGE NUMBER | +--+--+--+--+--+--+--+--+--+--+--+--+--+--+--+--+ 1 | HEADER CHECKSUM | +--+--+--+--+--+--+--+--+--+--+--+--+--+--+--+--+ 2 | A/R | +--+--+--+--+--+--+--+--+--+--+--+--+--+--+--+--+ 3 | 0|IL| D| E| PRI | TTL | RLY | RLEN | +--+--+--+--+--+--+--+--+--+--+--+--+--+--+--+--+ 4 | DESTINATION HOST ADDRESS | +--+--+--+--+--+--+--+--+--+--+--+--+--+--+--+--+ 5 | SOURCE HOST ADDRESS | +--+--+--+--+--+--+--+--+--+--+--+--+--+--+--+--+ 6 | PROTOCOL ID | +--+--+--+--+--+--+--+--+--+--+--+--+--+--+--+--+ | | 7-N : DATA : | | +--+--+--+--+--+--+--+--+--+--+--+--+--+--+--+--+ DATAGRAM MESSAGE Figure 1 0[0] Message Class. This bit identifies the message as a data message or a control message. 0 = Data Message 1 = Control Message 0[1] Loopback indicator. This bit allows the sender of a message to determine if its own messages are being looped back. The host and the WPS each use different settings of this bit for their transmissions. If a message arrives with the loopback bit set equal to itsEdmond [Page 7]RFC 1221 HAP2 April 1991 outgoing value, then the message has been looped. 0 = Sent by Host 1 = Sent by WPS 0[2-3] Go-Priority. In WPS-to-Host messages, this field provides advisory information concerning the lowest priority currently being accepted by the WPS. The host may optionally choose to provide similar priority information to the WPS. 0 = Low Priority 1 = Medium Priority 2 = High Priority 3 = (Reserved.) 0[4-6] Reserved. Must be zero. 0[7] Reserved. Must be zero. Formerly used for WPS diagnostic purposes. 0[8-15] Message Number. This field contains the identification of the message used by the acceptance/refusal (A/R) mechanism (when enabled). If the message number is zero, A/R is disabled for this specific message. See Section 5 for a detailed description of the A/R mechanism. 1[0-15] Header Checksum. The checksum is the 2's-complement of the 2's-complement sum of words 0-6 (excluding the checksum word itself). 2[0-15] Piggybacked A/R. This field may contain an acceptance/refusal word providing A/R status on traffic flowing in the opposite direction. Its inclusion may eliminate the need for a separate A/R control message (see Section 5). A value of zero for this word is used to indicate that no piggybacked A/R information is present. 3[0] Data Message Type. This bit identifies whether the message is a datagram message or a stream message. 0 = Datagram Message 1 = Stream Message 3[1] IL flag. Obsolete. Must be zero. (See Appendix B.)Edmond [Page 8]RFC 1221 HAP2 April 1991 3[2] Discard Flag. This flag allows a source host to instruct the network (including the destination host) what to do with the message when data errors are detected (assuming the header checksum is correct). 0 = Discard message if data errors detected. 1 = Don't discard message if data errors detected. The value of this flag, set by the source host, is passed on to the destination host. 3[3] Data Error Flag. This flag is used in conjunction with the Discard Flag to indicate to the destination host whether any data errors have been detected in the message prior to transmission over the destination's WPS-to-Host access link. It is used only if Discard Flag = 1. It should be set to zero by the source host. 0 = No Data Errors Detected 1 = Data Errors Detected 3[4-5] Priority. The source host uses this field to specify the priority with which the message should be handled within the network. 0 = Low Priority 1 = Medium Priority 2 = High Priority 3 = (Reserved.) The priority of each message is passed to the destination host by the destination WPS. 3[6-7] Time-to-Live Designator. The source host uses this field to specify the maximum time that a message should be allowed to exist within the network before being deleted. Elapsed time begins when the message has been received by the WPS from the source host (or is sent by a WPS agent) and is last checked when the message is queued for transmission out the I/O interface to the destination host. If a message is multicast, each copy is treated separately. 0 = 1 seconds 1 = 2 seconds 2 = 5 seconds 3 = 10 secondsEdmond [Page 9]RFC 1221 HAP2 April 1991 3[8-9] Reliability. The source host uses this field to specify the basic bit error rate requirement for the data portion of this message. The source WPS uses this field to determine the trunk circuit transmission parameters and forward error correction level required to provide that bit error rate. 0 = Low Reliability 1 = Medium-Low Reliability 2 = Medium-High Reliability 3 = High Reliability 3[10-15] Reliability Length. The source host uses this field to specify a portion of the user data which should be transmitted at the highest reliability level (lowest bit error rate). Both the HAP message header words and the first 2*<Reliability Length> octets of user data will be transmitted at high reliability while the remainder of the user data will be transmitted at whatever reliability level is specified in field 3[8- 9]. The reliability length mechanism gives the user the ability to transmit private header information (e.g., IP and TCP headers) at a higher reliability level than the remainder of the data. 4[0-15] Destination Host Address. This field contains the network logical address of the destination host. 5[0-15] Source Host Address. This field contains the network logical address of the source host. 6[0-15] Protocol ID. This field specifies the next higher level protocol. Protocol identifiers are assigned administratively, except 0 which is reserved, and are not part of this specification. See reference [10]. 7-N Data. This field contains up to 16,384 bits (2048 octets) of user data, and must be an even number of octets.4. Stream Messages Stream messages are the second message type provided by HAP, as described in Section 2. Streams provide guaranteed bandwidth between the source and destination(s), and provide the minimum delivery delay and delay variance available in the network. Streams are suitable for volatile traffic, such as speech, and for support of high duty cycle applications that require throughput guarantees.Edmond [Page 10]RFC 1221 HAP2 April 1991 Streams must be created before stream messages can flow from host to host. The protocol to accomplish stream creation is described in Section 6.1. Once established, a stream is allocated specific network resources, such as bandwidth. Within the bounds of its stream allocation, a host is permitted considerable flexibility in how it may use the stream. Although the time to live, reliability, and reliability length of each stream message is fixed at stream setup time, the destination logical address can vary from stream message to stream message. A host can, therefore, multiplex a variety of logical flows onto a single stream, as long as the stream was set up to reach all the destination hosts. The format of stream messages is described in Figure 2. 0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 +--+--+--+--+--+--+--+--+--+--+--+--+--+--+--+--+ 0 | 0|LB|GOPRI| 0 | MESSAGE NUMBER | +--+--+--+--+--+--+--+--+--+--+--+--+--+--+--+--+ 1 | HEADER CHECKSUM | +--+--+--+--+--+--+--+--+--+--+--+--+--+--+--+--+ 2 | A/R | +--+--+--+--+--+--+--+--+--+--+--+--+--+--+--+--+ 3 | 1|IL| D| E| PRI | HOST STREAM ID | +--+--+--+--+--+--+--+--+--+--+--+--+--+--+--+--+
⌨️ 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 + -