From 7c9b09bc511236c8cc5f6cea7623a8b98fedf302 Mon Sep 17 00:00:00 2001 From: Martin Lucina Date: Wed, 2 Jun 2010 18:36:34 +0200 Subject: Documentation: Flow control, zmq_socket(3) Mostly Flow control and additions to zmq_socket(3) Removed/changed lots of text regarding message queues More fixes for 2.0.7 changes --- doc/zmq.txt | 18 ++---- doc/zmq_getsockopt.txt | 23 ++++--- doc/zmq_poll.txt | 28 ++++---- doc/zmq_recv.txt | 22 +++---- doc/zmq_send.txt | 9 ++- doc/zmq_setsockopt.txt | 25 +++---- doc/zmq_socket.txt | 172 +++++++++++++++++++++++++++++++++++++++++-------- 7 files changed, 205 insertions(+), 92 deletions(-) diff --git a/doc/zmq.txt b/doc/zmq.txt index fd9f7ed..2d7f1c8 100644 --- a/doc/zmq.txt +++ b/doc/zmq.txt @@ -86,17 +86,9 @@ Message manipulation:: Sockets ~~~~~~~ -Standard sockets present a _synchronous_ interface to either connection-mode -reliable byte streams (SOCK_STREAM), or connection-less unreliable datagrams -(SOCK_DGRAM). In comparison, 0MQ sockets present an abstraction of a -asynchronous _message queue_, with the exact queueing semantics depending on -the socket type in use. See linkzmq:zmq_socket[3] for the socket types -provided. - -0MQ sockets being _asynchronous_ means that the timings of the physical -connection setup and teardown, reconnect and effective delivery are organized -by 0MQ itself, and that messages may be _queued_ in the event that a peer is -unavailable to receive them. +0MQ sockets present an abstraction of a asynchronous _message queue_, with the +exact queueing semantics depending on the socket type in use. See +linkzmq:zmq_socket[3] for the socket types provided. The following functions are provided to work with sockets: @@ -118,9 +110,7 @@ Sending and receiving messages:: linkzmq:zmq_send[3] linkzmq:zmq_recv[3] - -Input/output multiplexing -^^^^^^^^^^^^^^^^^^^^^^^^^ +.Input/output multiplexing 0MQ provides a mechanism for applications to multiplex input/output events over a set containing both 0MQ sockets and standard sockets. This mechanism mirrors the standard _poll()_ system call, and is described in detail in diff --git a/doc/zmq_getsockopt.txt b/doc/zmq_getsockopt.txt index 318d12e..f08748b 100644 --- a/doc/zmq_getsockopt.txt +++ b/doc/zmq_getsockopt.txt @@ -45,12 +45,16 @@ Applicable socket types:: all ZMQ_HWM: Retrieve high water mark ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ -The 'ZMQ_HWM' option shall retrieve the high water mark for the _message queue_ -associated with the specified 'socket'. The high water mark is a hard limit on -the number of outstanding messages in the queue; if this limit has been reached -the socket shall enter an "emergency" state and depending on the socket type, -0MQ shall take appropriate action such as blocking or dropping new messages -entering the queue. +The 'ZMQ_HWM' option shall retrieve the high water mark for the specified +'socket'. The high water mark is a hard limit on the maximum number of +outstanding messages 0MQ shall queue in memory for any single peer that the +specified 'socket' is communicating with. + +If this limit has been reached the socket shall enter an exceptional state and +depending on the socket type, 0MQ shall take appropriate action such as +blocking or dropping sent messages. Refer to the individual socket descriptions +in linkzmq:zmq_socket[3] for details on the exact action taken for each socket +type. The default 'ZMQ_HWM' value of zero means "no limit". @@ -63,10 +67,9 @@ Applicable socket types:: all ZMQ_SWAP: Retrieve disk offload size ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ The 'ZMQ_SWAP' option shall retrieve the disk offload (swap) size for the -_message queue_ associated with the specified 'socket'. A socket which has -'ZMQ_SWAP' set to a non-zero value may exceed it's high water mark; in this -case outstanding messages shall be offloaded to storage on disk rather than -held in memory. +specified 'socket'. A socket which has 'ZMQ_SWAP' set to a non-zero value may +exceed it's high water mark; in this case outstanding messages shall be +offloaded to storage on disk rather than held in memory. The value of 'ZMQ_SWAP' defines the maximum size of the swap space in bytes. diff --git a/doc/zmq_poll.txt b/doc/zmq_poll.txt index 9106b5b..d96af3e 100644 --- a/doc/zmq_poll.txt +++ b/doc/zmq_poll.txt @@ -54,18 +54,16 @@ The 'events' and 'revents' members of *zmq_pollitem_t* are bitmasks constructed by OR'ing a combination of the following event flags: *ZMQ_POLLIN*:: -For 0MQ sockets, at least one message may be dequeued from the underlying -_message queue_ associated with 'socket' without blocking. For standard sockets -this is equivalent to the 'POLLIN' flag of the _poll()_ system call and -generally means that at least one byte of data may be read from 'fd' without -blocking. +For 0MQ sockets, at least one message may be received from the 'socket' without +blocking. For standard sockets this is equivalent to the 'POLLIN' flag of the +_poll()_ system call and generally means that at least one byte of data may be +read from 'fd' without blocking. *ZMQ_POLLOUT*:: -For 0MQ sockets, at least one message may be queued on the underlying -_message queue_ associated with 'socket' without blocking. For standard sockets -this is equivalent to the 'POLLOUT' flag of the _poll()_ system call and -generally means that at least one byte of data may be written to 'fd' -without blocking. +For 0MQ sockets, at least one message may be sent to the 'socket' without +blocking. For standard sockets this is equivalent to the 'POLLOUT' flag of the +_poll()_ system call and generally means that at least one byte of data may be +written to 'fd' without blocking. *ZMQ_POLLERR*:: For standard sockets, this flag is passed through _zmq_poll()_ to the @@ -82,10 +80,12 @@ of those interfaces in ways not defined in this documentation. RETURN VALUE ------------ Upon successful completion, the _zmq_poll()_ function shall return the number -of *zmq_pollitem_t* structures with events signaled in 'revents' or `0` if the -'timeout' period has expired and no events have been signaled. Upon failure, -_zmq_poll()_ shall return `-1` and set 'errno' to one of the values defined -below. +of *zmq_pollitem_t* structures with events signaled in 'revents' or `0` if no +events have been signaled. Upon failure, _zmq_poll()_ shall return `-1` and set +'errno' to one of the values defined below. + +IMPORTANT: The _zmq_poll()_ function may return *before* the 'timeout' period +has expired even if no events have been signaled. ERRORS diff --git a/doc/zmq_recv.txt b/doc/zmq_recv.txt index 42df99a..79d3dc1 100644 --- a/doc/zmq_recv.txt +++ b/doc/zmq_recv.txt @@ -14,19 +14,17 @@ SYNOPSIS DESCRIPTION ----------- -The _zmq_recv()_ function shall dequeue a message from the underlying _message -queue_ associated with the socket referenced by the 'socket' argument and store -it in the message referenced by the 'msg' argument. Any content previously -stored in 'msg' shall be properly deallocated. If there are no messages -available to be dequeued from the underlying _message queue_ associated with -'socket' the _zmq_recv()_ function shall block until the request can be -satisfied. The 'flags' argument is a combination of the flags defined below: +The _zmq_recv()_ function shall receive a message from the socket referenced by +the 'socket' argument and store it in the message referenced by the 'msg' +argument. Any content previously stored in 'msg' shall be properly deallocated. +If there are no messages available on the specified 'socket' the _zmq_recv()_ +function shall block until the request can be satisfied. The 'flags' argument +is a combination of the flags defined below: *ZMQ_NOBLOCK*:: Specifies that the operation should be performed in non-blocking mode. If there -are no messages available to be dequeued from the underlying _message queue_ -associated with 'socket', the _zmq_recv()_ function shall fail with 'errno' set -to EAGAIN. +are no messages available on the specified 'socket', the _zmq_recv()_ function +shall fail with 'errno' set to EAGAIN. Multi-part messages @@ -75,7 +73,7 @@ EXAMPLE zmq_msg_t msg; int rc = zmq_msg_init (&msg); assert (rc == 0); -/* Block until a message is available to be dequeued from socket */ +/* Block until a message is available to be received from socket */ rc = zmq_recv (socket, &msg, 0); assert (rc == 0); ---- @@ -89,7 +87,7 @@ do { zmq_msg_t part; int rc = zmq_msg_init (&part); assert (rc == 0); - /* Block until a message is available to be dequeued from socket */ + /* Block until a message is available to be received from socket */ rc = zmq_recv (socket, &part, 0); assert (rc == 0); /* Determine if more message parts are to follow */ diff --git a/doc/zmq_send.txt b/doc/zmq_send.txt index 300f812..419e386 100644 --- a/doc/zmq_send.txt +++ b/doc/zmq_send.txt @@ -20,8 +20,8 @@ argument to be sent to the socket referenced by the 'socket' argument. The *ZMQ_NOBLOCK*:: Specifies that the operation should be performed in non-blocking mode. If the -message cannot be queued on the underlying _message queue_ associated with -'socket', the _zmq_send()_ function shall fail with 'errno' set to EAGAIN. +message cannot be queued on the 'socket', the _zmq_send()_ function shall fail +with 'errno' set to EAGAIN. *ZMQ_SNDMORE*:: Specifies that the message being sent is a multi-part message, and that further @@ -30,8 +30,7 @@ below for a detailed description. NOTE: A successful invocation of _zmq_send()_ does not indicate that the message has been transmitted to the network, only that it has been queued on -the _message queue_ associated with the socket and 0MQ has assumed -responsibility for the message. +the 'socket' and 0MQ has assumed responsibility for the message. Multi-part messages @@ -60,7 +59,7 @@ return `-1` and set 'errno' to one of the values defined below. ERRORS ------ *EAGAIN*:: -Non-blocking mode was requested and the message cannot be queued at the moment. +Non-blocking mode was requested and the message cannot be sent at the moment. *ENOTSUP*:: The _zmq_send()_ operation is not supported by this socket type. *EFSM*:: diff --git a/doc/zmq_setsockopt.txt b/doc/zmq_setsockopt.txt index 7721eff..93c5889 100644 --- a/doc/zmq_setsockopt.txt +++ b/doc/zmq_setsockopt.txt @@ -25,12 +25,16 @@ The following socket options can be set with the _zmq_setsockopt()_ function: ZMQ_HWM: Set high water mark ~~~~~~~~~~~~~~~~~~~~~~~~~~~~ -The 'ZMQ_HWM' option shall set the high water mark for the _message queue_ -associated with the specified 'socket'. The high water mark is a hard limit on -the number of outstanding messages in the queue; if this limit has been reached -the socket shall enter an "emergency" state and depending on the socket type, -0MQ shall take appropriate action such as blocking or dropping new messages -entering the queue. +The 'ZMQ_HWM' option shall set the high water mark for the specified 'socket'. +The high water mark is a hard limit on the maximum number of outstanding +messages 0MQ shall queue in memory for any single peer that the specified +'socket' is communicating with. + +If this limit has been reached the socket shall enter an exceptional state and +depending on the socket type, 0MQ shall take appropriate action such as +blocking or dropping sent messages. Refer to the individual socket descriptions +in linkzmq:zmq_socket[3] for details on the exact action taken for each socket +type. The default 'ZMQ_HWM' value of zero means "no limit". @@ -42,11 +46,10 @@ Applicable socket types:: all ZMQ_SWAP: Set disk offload size ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ -The 'ZMQ_SWAP' option shall set the disk offload (swap) size for the _message -queue_ associated with the specified 'socket'. A socket which has 'ZMQ_SWAP' -set to a non-zero value may exceed it's high water mark; in this case -outstanding messages shall be offloaded to storage on disk rather than held in -memory. +The 'ZMQ_SWAP' option shall set the disk offload (swap) size for the specified +'socket'. A socket which has 'ZMQ_SWAP' set to a non-zero value may exceed it's +high water mark; in this case outstanding messages shall be offloaded to +storage on disk rather than held in memory. The value of 'ZMQ_SWAP' defines the maximum size of the swap space in bytes. diff --git a/doc/zmq_socket.txt b/doc/zmq_socket.txt index 119ac31..23cc317 100644 --- a/doc/zmq_socket.txt +++ b/doc/zmq_socket.txt @@ -19,30 +19,89 @@ The 'zmq_socket()' function shall create a 0MQ socket within the specified argument specifies the socket type, which determines the semantics of communication over the socket. +The newly created socket is initially unbound, and not associated with any +endpoints. In order to establish a message flow a socket must first be +connected to at least one endpoint with linkzmq:zmq_connect[3], or at least one +endpoint must be created for accepting incoming connections with +linkzmq:zmq_bind[3]. + +.Key differences to conventional sockets +Generally speaking, conventional sockets present a _synchronous_ interface to +either connection-oriented reliable byte streams (SOCK_STREAM), or +connection-less unreliable datagrams (SOCK_DGRAM). In comparison, 0MQ sockets +present an abstraction of an asynchronous _message queue_, with the exact +queueing semantics depending on the socket type in use. Where conventional +sockets transfer streams of bytes or discrete datagrams, 0MQ sockets transfer +discrete _messages_. + +0MQ sockets being _asynchronous_ means that the timings of the physical +connection setup and teardown, reconnect and effective delivery are transparent +to the user and organized by 0MQ itself. Further, messages may be _queued_ in +the event that a peer is unavailable to receive them. + +Conventional sockets allow only strict one-to-one (two peers), many-to-one +(many clients, one server), or in some cases one-to-many (multicast) +relationships. With the exception of 'ZMQ_PAIR', 0MQ sockets may be connected +*to multiple endpoints* using _zmq_connect()_, while simultaneously accepting +incoming connections *from multiple endpoints* bound to the socket using +_zmq_bind()_, thus allowing many-to-many relationships. + +.Socket types The following sections present the socket types defined by 0MQ, grouped by the -general _messaging pattern_ built from related socket types. +general _messaging pattern_ which is built from related socket types. Request-reply pattern ~~~~~~~~~~~~~~~~~~~~~ -The request-reply pattern is used for sending requests from a _client_ to a -_service_, and receiving subsequent replies to each request sent. +The request-reply pattern is used for sending requests from a _client_ to one +or more instances of a _service_, and receiving subsequent replies to each +request sent. -Socket type:: 'ZMQ_REQ' -Compatible peer sockets:: 'ZMQ_REP' +ZMQ_REQ +^^^^^^^ A socket of type 'ZMQ_REQ' is used by a _client_ to send requests to and receive replies from a _service_. This socket type allows only an alternating sequence of _zmq_send(request)_ and subsequent _zmq_recv(reply)_ calls. Each -request sent is load-balanced among all connected _services_. +request sent is load-balanced among all _services_, and each reply received is +matched with the last issued request. -Socket type:: 'ZMQ_REP' -Compatible peer sockets:: 'ZMQ_REQ' +When a 'ZMQ_REQ' socket enters an exceptional state due to having reached the +high water mark for all _services_, or if there are no _services_ at all, then +any linkzmq:zmq_send[3] operations on the socket shall block until the +exceptional state ends or at least one _service_ becomes available for sending; +messages are not discarded. + +[horizontal] +.Summary of ZMQ_REQ characteristics +Compatible peer sockets:: 'ZMQ_REP' +Direction:: Bidirectional +Send/receive pattern:: Send, Receive, Send, Receive, ... +Outgoing routing strategy:: Load-balanced +Incoming routing strategy:: Last peer +ZMQ_HWM option action:: Block + +ZMQ_REP +^^^^^^^ A socket of type 'ZMQ_REP' is used by a _service_ to receive requests from and -send replies to a _client_. This socket type allows only an alternating +send replies to a _client_. This socket type allows only an alternating sequence of _zmq_recv(request)_ and subsequent _zmq_send(reply)_ calls. Each -reply is routed to the _client_ that issued the last received request. +request received is fair-queued from among all _clients_, and each reply sent +is routed to the _client_ that issued the last request. + +When a 'ZMQ_REP' socket enters an exceptional state due to having reached the +high water mark for a _client_, then any replies sent to the _client_ in +question shall be dropped until the exceptional state ends. + +[horizontal] +.Summary of ZMQ_REP characteristics +Compatible peer sockets:: 'ZMQ_REQ' +Direction:: Bidirectional +Send/receive pattern:: Receive, Send, Receive, Send, ... +Incoming routing strategy:: Fair-queued +Outgoing routing stratagy:: Last peer +ZMQ_HWM option action:: Drop Publish-subscribe pattern @@ -50,21 +109,44 @@ Publish-subscribe pattern The publish-subscribe pattern is used for one-to-many distribution of data from a single _publisher_ to multiple _subscribers_ in a fanout fashion. -Socket type:: 'ZMQ_PUB' -Compatible peer sockets:: 'ZMQ_SUB' +ZMQ_PUB +^^^^^^^ A socket of type 'ZMQ_PUB' is used by a _publisher_ to distribute data. Messages sent are distributed in a fanout fashion to all connected peers. -The _zmq_recv()_ function is not implemented for this socket type. +The linkzmq:zmq_recv[3] function is not implemented for this socket type. -Socket type:: 'ZMQ_SUB' -Compatible peer sockets:: 'ZMQ_PUB' +When a 'ZMQ_PUB' socket enters an exceptional state due to having reached the +high water mark for a _subscriber_, then any messages that would be sent to the +_subscriber_ in question shall instead be dropped until the exceptional state +ends. +[horizontal] +.Summary of ZMQ_PUB characteristics +Compatible peer sockets:: 'ZMQ_SUB' +Direction:: Unidirectional +Send/receive pattern:: Send only +Incoming routing strategy:: N/A +Outgoing routing strategy:: Fanout +ZMQ_HWM option action:: Drop + + +ZMQ_SUB +^^^^^^^ A socket of type 'ZMQ_SUB' is used by a _subscriber_ to subscribe to data distributed by a _publisher_. Initially a 'ZMQ_SUB' socket is not subscribed to -any messages, use the 'ZMQ_SUBSCRIBE' option of _zmq_setsockopt()_ to specify -which messages to subscribe to. The _zmq_send()_ function is not implemented -for this socket type. +any messages, use the 'ZMQ_SUBSCRIBE' option of linkzmq:zmq_setsockopt[3] to +specify which messages to subscribe to. The _zmq_send()_ function is not +implemented for this socket type. + +[horizontal] +.Summary of ZMQ_SUB characteristics +Compatible peer sockets:: 'ZMQ_PUB' +Direction:: Unidirectional +Send/receive pattern:: Receive only +Incoming routing strategy:: Fair-queued +Outgoing routing strategy:: N/A +ZMQ_HWM option action:: N/A Pipeline pattern @@ -74,38 +156,76 @@ a pipeline. Data always flows down the pipeline, and each stage of the pipeline is connected to at least one _node_. When a pipeline stage is connected to multiple _nodes_ data is load-balanced among all connected _nodes_. -Socket type:: 'ZMQ_DOWNSTREAM' -Compatible peer sockets:: 'ZMQ_UPSTREAM' +ZMQ_DOWNSTREAM +^^^^^^^^^^^^^^ A socket of type 'ZMQ_DOWNSTREAM' is used by a pipeline _node_ to send messages to downstream pipeline _nodes_. Messages are load-balanced to all connected downstream _nodes_. The _zmq_recv()_ function is not implemented for this socket type. -Socket type:: 'ZMQ_UPSTREAM' -Compatible peer sockets:: 'ZMQ_DOWNSTREAM' +When a 'ZMQ_DOWNSTREAM' socket enters an exceptional state due to having +reached the high water mark for all downstream _nodes_, or if there are no +downstream _nodes_ at all, then any linkzmq:zmq_send[3] operations on the +socket shall block until the exceptional state ends or at least one downstream +_node_ becomes available for sending; messages are not discarded. + +[horizontal] +.Summary of ZMQ_DOWNSTREAM characteristics +Compatible peer sockets:: 'ZMQ_UPSTREAM' +Direction:: Unidirectional +Send/receive pattern:: Send only +Incoming routing strategy:: N/A +Outgoing routing strategy:: Load-balanced +ZMQ_HWM option action:: Block + +ZMQ_UPSTREAM +^^^^^^^^^^^^ A socket of type 'ZMQ_UPSTREAM' is used by a pipeline _node_ to receive messages from upstream pipeline _nodes_. Messages are fair-queued from among all connected upstream _nodes_. The _zmq_send()_ function is not implemented for this socket type. +[horizontal] +.Summary of ZMQ_UPSTREAM characteristics +Compatible peer sockets:: 'ZMQ_DOWNSTREAM' +Direction:: Unidirectional +Send/receive pattern:: Receive only +Incoming routing strategy:: Fair-queued +Outgoing routing strategy:: N/A +ZMQ_HWM option action:: N/A + Exclusive pair pattern ~~~~~~~~~~~~~~~~~~~~~~ The exclusive pair is an advanced pattern used for communicating exclusively between two peers. -Socket type:: 'ZMQ_PAIR' -Compatible peer sockets:: 'ZMQ_PAIR' +ZMQ_PAIR +^^^^^^^^ A socket of type 'ZMQ_PAIR' can only be connected to a single peer at any one time. No message routing or filtering is performed on messages sent over a 'ZMQ_PAIR' socket. +When a 'ZMQ_PAIR' socket enters an exceptional state due to having reached the +high water mark for the connected peer, or if no peer is connected, then +any linkzmq:zmq_send[3] operations on the socket shall block until the peer +becomes available for sending; messages are not discarded. + NOTE: 'ZMQ_PAIR' sockets are experimental, and are currently missing several features such as auto-reconnection. +[horizontal] +.Summary of ZMQ_PAIR characteristics +Compatible peer sockets:: 'ZMQ_PAIR' +Direction:: Bidirectional +Send/receive pattern:: Unrestricted +Incoming routing strategy:: N/A +Outgoing routing strategy:: N/A +ZMQ_HWM option action:: Block + RETURN VALUE ------------ @@ -120,8 +240,7 @@ ERRORS The requested socket 'type' is invalid. *EMTHREAD*:: -The number of application threads using sockets within this 'context' has been -exceeded. See the 'app_threads' parameter of the _zmq_init()_ function. +The maximum number of sockets within this 'context' has been exceeded. SEE ALSO @@ -132,6 +251,7 @@ linkzmq:zmq_bind[3] linkzmq:zmq_connect[3] linkzmq:zmq_send[3] linkzmq:zmq_recv[3] +linkzmq:zmq[7] AUTHORS -- cgit v1.2.3