ViewVC Help
View File | Revision Log | Show Annotations | Download File
/cvs/cvsroot/AnyEvent-MP/MP.pm
(Generate patch)

Comparing cvsroot/AnyEvent-MP/MP.pm (file contents):
Revision 1.37 by root, Fri Aug 7 16:47:23 2009 UTC vs.
Revision 1.63 by root, Thu Aug 27 21:29:37 2009 UTC

8 8
9 $NODE # contains this node's noderef 9 $NODE # contains this node's noderef
10 NODE # returns this node's noderef 10 NODE # returns this node's noderef
11 NODE $port # returns the noderef of the port 11 NODE $port # returns the noderef of the port
12 12
13 $SELF # receiving/own port id in rcv callbacks
14
15 # initialise the node so it can send/receive messages
16 initialise_node;
17
18 # ports are message endpoints
19
20 # sending messages
13 snd $port, type => data...; 21 snd $port, type => data...;
22 snd $port, @msg;
23 snd @msg_with_first_element_being_a_port;
14 24
15 $SELF # receiving/own port id in rcv callbacks 25 # creating/using ports, the simple way
26 my $simple_port = port { my @msg = @_; 0 };
16 27
17 rcv $port, smartmatch => $cb->($port, @msg); 28 # creating/using ports, tagged message matching
18 29 my $port = port;
19 # examples:
20 rcv $port2, ping => sub { snd $_[0], "pong"; 0 }; 30 rcv $port, ping => sub { snd $_[0], "pong"; 0 };
21 rcv $port1, pong => sub { warn "pong received\n" }; 31 rcv $port, pong => sub { warn "pong received\n"; 0 };
22 snd $port2, ping => $port1;
23 32
24 # more, smarter, matches (_any_ is exported by this module) 33 # create a port on another node
25 rcv $port, [child_died => $pid] => sub { ... 34 my $port = spawn $node, $initfunc, @initdata;
26 rcv $port, [_any_, _any_, 3] => sub { .. $_[2] is 3
27 35
28 # monitoring 36 # monitoring
29 mon $port, $cb->(@msg) # callback is invoked on death 37 mon $port, $cb->(@msg) # callback is invoked on death
30 mon $port, $otherport # kill otherport on abnormal death 38 mon $port, $otherport # kill otherport on abnormal death
31 mon $port, $otherport, @msg # send message on death 39 mon $port, $otherport, @msg # send message on death
32 40
41=head1 CURRENT STATUS
42
43 AnyEvent::MP - stable API, should work
44 AnyEvent::MP::Intro - outdated
45 AnyEvent::MP::Kernel - WIP
46 AnyEvent::MP::Transport - mostly stable
47
48 stay tuned.
49
33=head1 DESCRIPTION 50=head1 DESCRIPTION
34 51
35This module (-family) implements a simple message passing framework. 52This module (-family) implements a simple message passing framework.
36 53
37Despite its simplicity, you can securely message other processes running 54Despite its simplicity, you can securely message other processes running
40For an introduction to this module family, see the L<AnyEvent::MP::Intro> 57For an introduction to this module family, see the L<AnyEvent::MP::Intro>
41manual page. 58manual page.
42 59
43At the moment, this module family is severly broken and underdocumented, 60At the moment, this module family is severly broken and underdocumented,
44so do not use. This was uploaded mainly to reserve the CPAN namespace - 61so do not use. This was uploaded mainly to reserve the CPAN namespace -
45stay tuned! The basic API should be finished, however. 62stay tuned!
46 63
47=head1 CONCEPTS 64=head1 CONCEPTS
48 65
49=over 4 66=over 4
50 67
51=item port 68=item port
52 69
53A port is something you can send messages to (with the C<snd> function). 70A port is something you can send messages to (with the C<snd> function).
54 71
55Some ports allow you to register C<rcv> handlers that can match specific 72Ports allow you to register C<rcv> handlers that can match all or just
56messages. All C<rcv> handlers will receive messages they match, messages 73some messages. Messages will not be queued.
57will not be queued.
58 74
59=item port id - C<noderef#portname> 75=item port ID - C<noderef#portname>
60 76
61A port id is normaly the concatenation of a noderef, a hash-mark (C<#>) as 77A port ID is the concatenation of a noderef, a hash-mark (C<#>) as
62separator, and a port name (a printable string of unspecified format). An 78separator, and a port name (a printable string of unspecified format). An
63exception is the the node port, whose ID is identical to its node 79exception is the the node port, whose ID is identical to its node
64reference. 80reference.
65 81
66=item node 82=item node
67 83
68A node is a single process containing at least one port - the node 84A node is a single process containing at least one port - the node port,
69port. You can send messages to node ports to find existing ports or to 85which provides nodes to manage each other remotely, and to create new
70create new ports, among other things. 86ports.
71 87
72Nodes are either private (single-process only), slaves (connected to a 88Nodes are either private (single-process only), slaves (can only talk to
73master node only) or public nodes (connectable from unrelated nodes). 89public nodes, but do not need an open port) or public nodes (connectable
90from any other node).
74 91
75=item noderef - C<host:port,host:port...>, C<id@noderef>, C<id> 92=item node ID - C<[a-za-Z0-9_\-.:]+>
76 93
77A node reference is a string that either simply identifies the node (for 94A node ID is a string that either simply identifies the node (for
78private and slave nodes), or contains a recipe on how to reach a given 95private and slave nodes), or contains a recipe on how to reach a given
79node (for public nodes). 96node (for public nodes).
80 97
81This recipe is simply a comma-separated list of C<address:port> pairs (for 98This recipe is simply a comma-separated list of C<address:port> pairs (for
82TCP/IP, other protocols might look different). 99TCP/IP, other protocols might look different).
95 112
96=cut 113=cut
97 114
98package AnyEvent::MP; 115package AnyEvent::MP;
99 116
100use AnyEvent::MP::Base; 117use AnyEvent::MP::Kernel;
101 118
102use common::sense; 119use common::sense;
103 120
104use Carp (); 121use Carp ();
105 122
106use AE (); 123use AE ();
107 124
108use base "Exporter"; 125use base "Exporter";
109 126
110our $VERSION = '0.1'; 127our $VERSION = $AnyEvent::MP::Kernel::VERSION;
128
111our @EXPORT = qw( 129our @EXPORT = qw(
112 NODE $NODE *SELF node_of _any_ 130 NODE $NODE *SELF node_of after
113 resolve_node initialise_node 131 resolve_node initialise_node
114 snd rcv mon kil reg psub 132 snd rcv mon mon_guard kil reg psub spawn
115 port 133 port
116); 134);
117 135
118our $SELF; 136our $SELF;
119 137
123 kil $SELF, die => $msg; 141 kil $SELF, die => $msg;
124} 142}
125 143
126=item $thisnode = NODE / $NODE 144=item $thisnode = NODE / $NODE
127 145
128The C<NODE> function returns, and the C<$NODE> variable contains 146The C<NODE> function returns, and the C<$NODE> variable contains the
129the noderef of the local node. The value is initialised by a call 147node id of the local node. The value is initialised by a call to
130to C<become_public> or C<become_slave>, after which all local port 148C<initialise_node>.
131identifiers become invalid.
132 149
133=item $noderef = node_of $port 150=item $nodeid = node_of $port
134 151
135Extracts and returns the noderef from a portid or a noderef. 152Extracts and returns the noderef from a port ID or a node ID.
136 153
137=item initialise_node $noderef, $seednode, $seednode... 154=item initialise_node $profile_name
138
139=item initialise_node "slave/", $master, $master...
140 155
141Before a node can talk to other nodes on the network it has to initialise 156Before a node can talk to other nodes on the network it has to initialise
142itself - the minimum a node needs to know is it's own name, and optionally 157itself - the minimum a node needs to know is it's own name, and optionally
143it should know the noderefs of some other nodes in the network. 158it should know the noderefs of some other nodes in the network.
144 159
145This function initialises a node - it must be called exactly once (or 160This function initialises a node - it must be called exactly once (or
146never) before calling other AnyEvent::MP functions. 161never) before calling other AnyEvent::MP functions.
147 162
148All arguments are noderefs, which can be either resolved or unresolved. 163All arguments (optionally except for the first) are noderefs, which can be
164either resolved or unresolved.
165
166The first argument will be looked up in the configuration database first
167(if it is C<undef> then the current nodename will be used instead) to find
168the relevant configuration profile (see L<aemp>). If none is found then
169the default configuration is used. The configuration supplies additional
170seed/master nodes and can override the actual noderef.
149 171
150There are two types of networked nodes, public nodes and slave nodes: 172There are two types of networked nodes, public nodes and slave nodes:
151 173
152=over 4 174=over 4
153 175
154=item public nodes 176=item public nodes
155 177
156For public nodes, C<$noderef> must either be a (possibly unresolved) 178For public nodes, C<$noderef> (supplied either directly to
157noderef, in which case it will be resolved, or C<undef> (or missing), in 179C<initialise_node> or indirectly via a profile or the nodename) must be a
158which case the noderef will be guessed. 180noderef (possibly unresolved, in which case it will be resolved).
159 181
160Afterwards, the node will bind itself on all endpoints and try to connect 182After resolving, the node will bind itself on all endpoints.
161to all additional C<$seednodes> that are specified. Seednodes are optional
162and can be used to quickly bootstrap the node into an existing network.
163 183
164=item slave nodes 184=item slave nodes
165 185
166When the C<$noderef> is the special string C<slave/>, then the node will 186When the C<$noderef> (either as given or overriden by the config file)
187is the special string C<slave/>, then the node will become a slave
167become a slave node. Slave nodes cannot be contacted from outside and will 188node. Slave nodes cannot be contacted from outside, and cannot talk to
168route most of their traffic to the master node that they attach to. 189each other (at least in this version of AnyEvent::MP).
169 190
170At least one additional noderef is required: The node will try to connect 191Slave nodes work by creating connections to all public nodes, using the
171to all of them and will become a slave attached to the first node it can 192L<AnyEvent::MP::Global> service.
172successfully connect to.
173 193
174=back 194=back
175 195
176This function will block until all nodes have been resolved and, for slave 196After initialising itself, the node will connect to all additional
177nodes, until it has successfully established a connection to a master 197C<$seednodes> that are specified diretcly or via a profile. Seednodes are
178server. 198optional and can be used to quickly bootstrap the node into an existing
199network.
179 200
201All the seednodes will also be specially marked to automatically retry
202connecting to them indefinitely, so make sure that seednodes are really
203reliable and up (this might also change in the future).
204
180Example: become a public node listening on the default node. 205Example: become a public node listening on the guessed noderef, or the one
206specified via C<aemp> for the current node. This should be the most common
207form of invocation for "daemon"-type nodes.
181 208
182 initialise_node; 209 initialise_node;
210
211Example: become a slave node to any of the the seednodes specified via
212C<aemp>. This form is often used for commandline clients.
213
214 initialise_node "slave/";
183 215
184Example: become a public node, and try to contact some well-known master 216Example: become a public node, and try to contact some well-known master
185servers to become part of the network. 217servers to become part of the network.
186 218
187 initialise_node undef, "master1", "master2"; 219 initialise_node undef, "master1", "master2";
190 222
191 initialise_node 4041; 223 initialise_node 4041;
192 224
193Example: become a public node, only visible on localhost port 4044. 225Example: become a public node, only visible on localhost port 4044.
194 226
195 initialise_node "locahost:4044"; 227 initialise_node "localhost:4044";
196
197Example: become a slave node to any of the specified master servers.
198
199 initialise_node "slave/", "master1", "192.168.13.17", "mp.example.net";
200 228
201=item $cv = resolve_node $noderef 229=item $cv = resolve_node $noderef
202 230
203Takes an unresolved node reference that may contain hostnames and 231Takes an unresolved node reference that may contain hostnames and
204abbreviated IDs, resolves all of them and returns a resolved node 232abbreviated IDs, resolves all of them and returns a resolved node
241=item snd $port, type => @data 269=item snd $port, type => @data
242 270
243=item snd $port, @msg 271=item snd $port, @msg
244 272
245Send the given message to the given port ID, which can identify either 273Send the given message to the given port ID, which can identify either
246a local or a remote port, and can be either a string or soemthignt hat 274a local or a remote port, and must be a port ID.
247stringifies a sa port ID (such as a port object :).
248 275
249While the message can be about anything, it is highly recommended to use a 276While the message can be about anything, it is highly recommended to use a
250string as first element (a portid, or some word that indicates a request 277string as first element (a port ID, or some word that indicates a request
251type etc.). 278type etc.).
252 279
253The message data effectively becomes read-only after a call to this 280The message data effectively becomes read-only after a call to this
254function: modifying any argument is not allowed and can cause many 281function: modifying any argument is not allowed and can cause many
255problems. 282problems.
260that Storable can serialise and deserialise is allowed, and for the local 287that Storable can serialise and deserialise is allowed, and for the local
261node, anything can be passed. 288node, anything can be passed.
262 289
263=item $local_port = port 290=item $local_port = port
264 291
265Create a new local port object that can be used either as a pattern 292Create a new local port object and returns its port ID. Initially it has
266matching port ("full port") or a single-callback port ("miniport"), 293no callbacks set and will throw an error when it receives messages.
267depending on how C<rcv> callbacks are bound to the object.
268 294
269=item $port = port { my @msg = @_; $finished } 295=item $local_port = port { my @msg = @_ }
270 296
271Creates a "miniport", that is, a very lightweight port without any pattern 297Creates a new local port, and returns its ID. Semantically the same as
272matching behind it, and returns its ID. Semantically the same as creating
273a port and calling C<rcv $port, $callback> on it. 298creating a port and calling C<rcv $port, $callback> on it.
274 299
275The block will be called for every message received on the port. When the 300The block will be called for every message received on the port, with the
276callback returns a true value its job is considered "done" and the port 301global variable C<$SELF> set to the port ID. Runtime errors will cause the
277will be destroyed. Otherwise it will stay alive. 302port to be C<kil>ed. The message will be passed as-is, no extra argument
303(i.e. no port ID) will be passed to the callback.
278 304
279The message will be passed as-is, no extra argument (i.e. no port id) will 305If you want to stop/destroy the port, simply C<kil> it:
280be passed to the callback.
281 306
282If you need the local port id in the callback, this works nicely: 307 my $port = port {
283 308 my @msg = @_;
284 my $port; $port = port { 309 ...
285 snd $otherport, reply => $port; 310 kil $SELF;
286 }; 311 };
287 312
288=cut 313=cut
289 314
290sub rcv($@); 315sub rcv($@);
316
317sub _kilme {
318 die "received message on port without callback";
319}
291 320
292sub port(;&) { 321sub port(;&) {
293 my $id = "$UNIQ." . $ID++; 322 my $id = "$UNIQ." . $ID++;
294 my $port = "$NODE#$id"; 323 my $port = "$NODE#$id";
295 324
296 if (@_) { 325 rcv $port, shift || \&_kilme;
297 rcv $port, shift;
298 } else {
299 $PORT{$id} = sub { }; # nop
300 }
301 326
302 $port 327 $port
303} 328}
304 329
305=item reg $port, $name
306
307=item reg $name
308
309Registers the given port (or C<$SELF><<< if missing) under the name
310C<$name>. If the name already exists it is replaced.
311
312A port can only be registered under one well known name.
313
314A port automatically becomes unregistered when it is killed.
315
316=cut
317
318sub reg(@) {
319 my $port = @_ > 1 ? shift : $SELF || Carp::croak 'reg: called with one argument only, but $SELF not set,';
320
321 $REG{$_[0]} = $port;
322}
323
324=item rcv $port, $callback->(@msg) 330=item rcv $local_port, $callback->(@msg)
325 331
326Replaces the callback on the specified miniport (after converting it to 332Replaces the default callback on the specified port. There is no way to
327one if required). 333remove the default callback: use C<sub { }> to disable it, or better
328 334C<kil> the port when it is no longer needed.
329=item rcv $port, tagstring => $callback->(@msg), ...
330
331=item rcv $port, $smartmatch => $callback->(@msg), ...
332
333=item rcv $port, [$smartmatch...] => $callback->(@msg), ...
334
335Register callbacks to be called on matching messages on the given full
336port (after converting it to one if required) and return the port.
337
338The callback has to return a true value when its work is done, after
339which is will be removed, or a false value in which case it will stay
340registered.
341 335
342The global C<$SELF> (exported by this module) contains C<$port> while 336The global C<$SELF> (exported by this module) contains C<$port> while
343executing the callback. 337executing the callback. Runtime errors during callback execution will
338result in the port being C<kil>ed.
344 339
345Runtime errors wdurign callback execution will result in the port being 340The default callback received all messages not matched by a more specific
346C<kil>ed. 341C<tag> match.
347 342
348If the match is an array reference, then it will be matched against the 343=item rcv $local_port, tag => $callback->(@msg_without_tag), ...
349first elements of the message, otherwise only the first element is being
350matched.
351 344
352Any element in the match that is specified as C<_any_> (a function 345Register (or replace) callbacks to be called on messages starting with the
353exported by this module) matches any single element of the message. 346given tag on the given port (and return the port), or unregister it (when
347C<$callback> is C<$undef> or missing). There can only be one callback
348registered for each tag.
354 349
355While not required, it is highly recommended that the first matching 350The original message will be passed to the callback, after the first
356element is a string identifying the message. The one-string-only match is 351element (the tag) has been removed. The callback will use the same
357also the most efficient match (by far). 352environment as the default callback (see above).
358 353
359Example: create a port and bind receivers on it in one go. 354Example: create a port and bind receivers on it in one go.
360 355
361 my $port = rcv port, 356 my $port = rcv port,
362 msg1 => sub { ...; 0 }, 357 msg1 => sub { ... },
363 msg2 => sub { ...; 0 }, 358 msg2 => sub { ... },
364 ; 359 ;
365 360
366Example: create a port, bind receivers and send it in a message elsewhere 361Example: create a port, bind receivers and send it in a message elsewhere
367in one go: 362in one go:
368 363
369 snd $otherport, reply => 364 snd $otherport, reply =>
370 rcv port, 365 rcv port,
371 msg1 => sub { ...; 0 }, 366 msg1 => sub { ... },
372 ... 367 ...
373 ; 368 ;
369
370Example: temporarily register a rcv callback for a tag matching some port
371(e.g. for a rpc reply) and unregister it after a message was received.
372
373 rcv $port, $otherport => sub {
374 my @reply = @_;
375
376 rcv $SELF, $otherport;
377 };
374 378
375=cut 379=cut
376 380
377sub rcv($@) { 381sub rcv($@) {
378 my $port = shift; 382 my $port = shift;
379 my ($noderef, $portid) = split /#/, $port, 2; 383 my ($noderef, $portid) = split /#/, $port, 2;
380 384
381 ($NODE{$noderef} || add_node $noderef) == $NODE{""} 385 $NODE{$noderef} == $NODE{""}
382 or Carp::croak "$port: rcv can only be called on local ports, caught"; 386 or Carp::croak "$port: rcv can only be called on local ports, caught";
383 387
384 if (@_ == 1) { 388 while (@_) {
389 if (ref $_[0]) {
390 if (my $self = $PORT_DATA{$portid}) {
391 "AnyEvent::MP::Port" eq ref $self
392 or Carp::croak "$port: rcv can only be called on message matching ports, caught";
393
394 $self->[2] = shift;
395 } else {
385 my $cb = shift; 396 my $cb = shift;
386 delete $PORT_DATA{$portid};
387 $PORT{$portid} = sub { 397 $PORT{$portid} = sub {
388 local $SELF = $port; 398 local $SELF = $port;
389 eval { 399 eval { &$cb }; _self_die if $@;
390 &$cb 400 };
391 and kil $port;
392 }; 401 }
393 _self_die if $@; 402 } elsif (defined $_[0]) {
394 };
395 } else {
396 my $self = $PORT_DATA{$portid} ||= do { 403 my $self = $PORT_DATA{$portid} ||= do {
397 my $self = bless { 404 my $self = bless [$PORT{$port} || sub { }, { }, $port], "AnyEvent::MP::Port";
398 id => $port,
399 }, "AnyEvent::MP::Port";
400 405
401 $PORT{$portid} = sub { 406 $PORT{$portid} = sub {
402 local $SELF = $port; 407 local $SELF = $port;
403 408
404 eval {
405 for (@{ $self->{rc0}{$_[0]} }) { 409 if (my $cb = $self->[1]{$_[0]}) {
406 $_ && &{$_->[0]} 410 shift;
407 && undef $_; 411 eval { &$cb }; _self_die if $@;
408 } 412 } else {
409
410 for (@{ $self->{rcv}{$_[0]} }) {
411 $_ && [@_[1 .. @{$_->[1]}]] ~~ $_->[1]
412 && &{$_->[0]} 413 &{ $self->[0] };
413 && undef $_;
414 }
415
416 for (@{ $self->{any} }) {
417 $_ && [@_[0 .. $#{$_->[1]}]] ~~ $_->[1]
418 && &{$_->[0]}
419 && undef $_;
420 } 414 }
421 }; 415 };
422 _self_die if $@; 416
417 $self
423 }; 418 };
424 419
425 $self
426 };
427
428 "AnyEvent::MP::Port" eq ref $self 420 "AnyEvent::MP::Port" eq ref $self
429 or Carp::croak "$port: rcv can only be called on message matching ports, caught"; 421 or Carp::croak "$port: rcv can only be called on message matching ports, caught";
430 422
431 while (@_) {
432 my ($match, $cb) = splice @_, 0, 2; 423 my ($tag, $cb) = splice @_, 0, 2;
433 424
434 if (!ref $match) { 425 if (defined $cb) {
435 push @{ $self->{rc0}{$match} }, [$cb]; 426 $self->[1]{$tag} = $cb;
436 } elsif (("ARRAY" eq ref $match && !ref $match->[0])) {
437 my ($type, @match) = @$match;
438 @match
439 ? push @{ $self->{rcv}{$match->[0]} }, [$cb, \@match]
440 : push @{ $self->{rc0}{$match->[0]} }, [$cb];
441 } else { 427 } else {
442 push @{ $self->{any} }, [$cb, $match]; 428 delete $self->[1]{$tag};
443 } 429 }
444 } 430 }
445 } 431 }
446 432
447 $port 433 $port
491 477
492=item $guard = mon $port 478=item $guard = mon $port
493 479
494=item $guard = mon $port, $rcvport, @msg 480=item $guard = mon $port, $rcvport, @msg
495 481
496Monitor the given port and do something when the port is killed, and 482Monitor the given port and do something when the port is killed or
497optionally return a guard that can be used to stop monitoring again. 483messages to it were lost, and optionally return a guard that can be used
484to stop monitoring again.
485
486C<mon> effectively guarantees that, in the absence of hardware failures,
487that after starting the monitor, either all messages sent to the port
488will arrive, or the monitoring action will be invoked after possible
489message loss has been detected. No messages will be lost "in between"
490(after the first lost message no further messages will be received by the
491port). After the monitoring action was invoked, further messages might get
492delivered again.
493
494Note that monitoring-actions are one-shot: once released, they are removed
495and will not trigger again.
498 496
499In the first form (callback), the callback is simply called with any 497In the first form (callback), the callback is simply called with any
500number of C<@reason> elements (no @reason means that the port was deleted 498number of C<@reason> elements (no @reason means that the port was deleted
501"normally"). Note also that I<< the callback B<must> never die >>, so use 499"normally"). Note also that I<< the callback B<must> never die >>, so use
502C<eval> if unsure. 500C<eval> if unsure.
503 501
504In the second form (another port given), the other port (C<$rcvport) 502In the second form (another port given), the other port (C<$rcvport>)
505will be C<kil>'ed with C<@reason>, iff a @reason was specified, i.e. on 503will be C<kil>'ed with C<@reason>, iff a @reason was specified, i.e. on
506"normal" kils nothing happens, while under all other conditions, the other 504"normal" kils nothing happens, while under all other conditions, the other
507port is killed with the same reason. 505port is killed with the same reason.
508 506
509The third form (kill self) is the same as the second form, except that 507The third form (kill self) is the same as the second form, except that
536sub mon { 534sub mon {
537 my ($noderef, $port) = split /#/, shift, 2; 535 my ($noderef, $port) = split /#/, shift, 2;
538 536
539 my $node = $NODE{$noderef} || add_node $noderef; 537 my $node = $NODE{$noderef} || add_node $noderef;
540 538
541 my $cb = @_ ? $_[0] : $SELF || Carp::croak 'mon: called with one argument only, but $SELF not set,'; 539 my $cb = @_ ? shift : $SELF || Carp::croak 'mon: called with one argument only, but $SELF not set,';
542 540
543 unless (ref $cb) { 541 unless (ref $cb) {
544 if (@_) { 542 if (@_) {
545 # send a kill info message 543 # send a kill info message
546 my (@msg) = @_; 544 my (@msg) = ($cb, @_);
547 $cb = sub { snd @msg, @_ }; 545 $cb = sub { snd @msg, @_ };
548 } else { 546 } else {
549 # simply kill other port 547 # simply kill other port
550 my $port = $cb; 548 my $port = $cb;
551 $cb = sub { kil $port, @_ if @_ }; 549 $cb = sub { kil $port, @_ if @_ };
598will be reported as reason C<< die => $@ >>. 596will be reported as reason C<< die => $@ >>.
599 597
600Transport/communication errors are reported as C<< transport_error => 598Transport/communication errors are reported as C<< transport_error =>
601$message >>. 599$message >>.
602 600
603=back
604
605=head1 NODE MESSAGES
606
607Nodes understand the following messages sent to them. Many of them take
608arguments called C<@reply>, which will simply be used to compose a reply
609message - C<$reply[0]> is the port to reply to, C<$reply[1]> the type and
610the remaining arguments are simply the message data.
611
612While other messages exist, they are not public and subject to change.
613
614=over 4
615
616=cut 601=cut
617 602
618=item lookup => $name, @reply 603=item $port = spawn $node, $initfunc[, @initdata]
619 604
620Replies with the port ID of the specified well-known port, or C<undef>. 605Creates a port on the node C<$node> (which can also be a port ID, in which
606case it's the node where that port resides).
621 607
622=item devnull => ... 608The port ID of the newly created port is return immediately, and it is
609permissible to immediately start sending messages or monitor the port.
623 610
624Generic data sink/CPU heat conversion. 611After the port has been created, the init function is
612called. This function must be a fully-qualified function name
613(e.g. C<MyApp::Chat::Server::init>). To specify a function in the main
614program, use C<::name>.
625 615
626=item relay => $port, @msg 616If the function doesn't exist, then the node tries to C<require>
617the package, then the package above the package and so on (e.g.
618C<MyApp::Chat::Server>, C<MyApp::Chat>, C<MyApp>) until the function
619exists or it runs out of package names.
627 620
628Simply forwards the message to the given port. 621The init function is then called with the newly-created port as context
622object (C<$SELF>) and the C<@initdata> values as arguments.
629 623
630=item eval => $string[ @reply] 624A common idiom is to pass your own port, monitor the spawned port, and
625in the init function, monitor the original port. This two-way monitoring
626ensures that both ports get cleaned up when there is a problem.
631 627
632Evaluates the given string. If C<@reply> is given, then a message of the 628Example: spawn a chat server port on C<$othernode>.
633form C<@reply, $@, @evalres> is sent.
634 629
635Example: crash another node. 630 # this node, executed from within a port context:
631 my $server = spawn $othernode, "MyApp::Chat::Server::connect", $SELF;
632 mon $server;
636 633
637 snd $othernode, eval => "exit"; 634 # init function on C<$othernode>
635 sub connect {
636 my ($srcport) = @_;
638 637
639=item time => @reply 638 mon $srcport;
640 639
641Replies the the current node time to C<@reply>. 640 rcv $SELF, sub {
641 ...
642 };
643 }
642 644
643Example: tell the current node to send the current time to C<$myport> in a 645=cut
644C<timereply> message.
645 646
646 snd $NODE, time => $myport, timereply => 1, 2; 647sub _spawn {
647 # => snd $myport, timereply => 1, 2, <time> 648 my $port = shift;
649 my $init = shift;
650
651 local $SELF = "$NODE#$port";
652 eval {
653 &{ load_func $init }
654 };
655 _self_die if $@;
656}
657
658sub spawn(@) {
659 my ($noderef, undef) = split /#/, shift, 2;
660
661 my $id = "$RUNIQ." . $ID++;
662
663 $_[0] =~ /::/
664 or Carp::croak "spawn init function must be a fully-qualified name, caught";
665
666 snd_to_func $noderef, "AnyEvent::MP::_spawn" => $id, @_;
667
668 "$noderef#$id"
669}
670
671=item after $timeout, @msg
672
673=item after $timeout, $callback
674
675Either sends the given message, or call the given callback, after the
676specified number of seconds.
677
678This is simply a utility function that come sin handy at times.
679
680=cut
681
682sub after($@) {
683 my ($timeout, @action) = @_;
684
685 my $t; $t = AE::timer $timeout, 0, sub {
686 undef $t;
687 ref $action[0]
688 ? $action[0]()
689 : snd @action;
690 };
691}
648 692
649=back 693=back
650 694
651=head1 AnyEvent::MP vs. Distributed Erlang 695=head1 AnyEvent::MP vs. Distributed Erlang
652 696
671convenience functionality. 715convenience functionality.
672 716
673This means that AEMP requires a less tightly controlled environment at the 717This means that AEMP requires a less tightly controlled environment at the
674cost of longer node references and a slightly higher management overhead. 718cost of longer node references and a slightly higher management overhead.
675 719
720=item * Erlang has a "remote ports are like local ports" philosophy, AEMP
721uses "local ports are like remote ports".
722
723The failure modes for local ports are quite different (runtime errors
724only) then for remote ports - when a local port dies, you I<know> it dies,
725when a connection to another node dies, you know nothing about the other
726port.
727
728Erlang pretends remote ports are as reliable as local ports, even when
729they are not.
730
731AEMP encourages a "treat remote ports differently" philosophy, with local
732ports being the special case/exception, where transport errors cannot
733occur.
734
676=item * Erlang uses processes and a mailbox, AEMP does not queue. 735=item * Erlang uses processes and a mailbox, AEMP does not queue.
677 736
678Erlang uses processes that selctively receive messages, and therefore 737Erlang uses processes that selectively receive messages, and therefore
679needs a queue. AEMP is event based, queuing messages would serve no useful 738needs a queue. AEMP is event based, queuing messages would serve no
680purpose. 739useful purpose. For the same reason the pattern-matching abilities of
740AnyEvent::MP are more limited, as there is little need to be able to
741filter messages without dequeing them.
681 742
682(But see L<Coro::MP> for a more Erlang-like process model on top of AEMP). 743(But see L<Coro::MP> for a more Erlang-like process model on top of AEMP).
683 744
684=item * Erlang sends are synchronous, AEMP sends are asynchronous. 745=item * Erlang sends are synchronous, AEMP sends are asynchronous.
685 746
686Sending messages in Erlang is synchronous and blocks the process. AEMP 747Sending messages in Erlang is synchronous and blocks the process (and
687sends are immediate, connection establishment is handled in the 748so does not need a queue that can overflow). AEMP sends are immediate,
688background. 749connection establishment is handled in the background.
689 750
690=item * Erlang can silently lose messages, AEMP cannot. 751=item * Erlang suffers from silent message loss, AEMP does not.
691 752
692Erlang makes few guarantees on messages delivery - messages can get lost 753Erlang makes few guarantees on messages delivery - messages can get lost
693without any of the processes realising it (i.e. you send messages a, b, 754without any of the processes realising it (i.e. you send messages a, b,
694and c, and the other side only receives messages a and c). 755and c, and the other side only receives messages a and c).
695 756
707eventually be killed - it cannot happen that a node detects a port as dead 768eventually be killed - it cannot happen that a node detects a port as dead
708and then later sends messages to it, finding it is still alive. 769and then later sends messages to it, finding it is still alive.
709 770
710=item * Erlang can send messages to the wrong port, AEMP does not. 771=item * Erlang can send messages to the wrong port, AEMP does not.
711 772
712In Erlang it is quite possible that a node that restarts reuses a process 773In Erlang it is quite likely that a node that restarts reuses a process ID
713ID known to other nodes for a completely different process, causing 774known to other nodes for a completely different process, causing messages
714messages destined for that process to end up in an unrelated process. 775destined for that process to end up in an unrelated process.
715 776
716AEMP never reuses port IDs, so old messages or old port IDs floating 777AEMP never reuses port IDs, so old messages or old port IDs floating
717around in the network will not be sent to an unrelated port. 778around in the network will not be sent to an unrelated port.
718 779
719=item * Erlang uses unprotected connections, AEMP uses secure 780=item * Erlang uses unprotected connections, AEMP uses secure
755This also saves round-trips and avoids sending messages to the wrong port 816This also saves round-trips and avoids sending messages to the wrong port
756(hard to do in Erlang). 817(hard to do in Erlang).
757 818
758=back 819=back
759 820
821=head1 RATIONALE
822
823=over 4
824
825=item Why strings for ports and noderefs, why not objects?
826
827We considered "objects", but found that the actual number of methods
828thatc an be called are very low. Since port IDs and noderefs travel over
829the network frequently, the serialising/deserialising would add lots of
830overhead, as well as having to keep a proxy object.
831
832Strings can easily be printed, easily serialised etc. and need no special
833procedures to be "valid".
834
835And a a miniport consists of a single closure stored in a global hash - it
836can't become much cheaper.
837
838=item Why favour JSON, why not real serialising format such as Storable?
839
840In fact, any AnyEvent::MP node will happily accept Storable as framing
841format, but currently there is no way to make a node use Storable by
842default.
843
844The default framing protocol is JSON because a) JSON::XS is many times
845faster for small messages and b) most importantly, after years of
846experience we found that object serialisation is causing more problems
847than it gains: Just like function calls, objects simply do not travel
848easily over the network, mostly because they will always be a copy, so you
849always have to re-think your design.
850
851Keeping your messages simple, concentrating on data structures rather than
852objects, will keep your messages clean, tidy and efficient.
853
854=back
855
760=head1 SEE ALSO 856=head1 SEE ALSO
761 857
762L<AnyEvent>. 858L<AnyEvent>.
763 859
764=head1 AUTHOR 860=head1 AUTHOR

Diff Legend

Removed lines
+ Added lines
< Changed lines
> Changed lines