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

Comparing AnyEvent-MP/MP.pm (file contents):
Revision 1.7 by root, Sat Aug 1 15:04:30 2009 UTC vs.
Revision 1.54 by root, Fri Aug 14 16:15:37 2009 UTC

4 4
5=head1 SYNOPSIS 5=head1 SYNOPSIS
6 6
7 use AnyEvent::MP; 7 use AnyEvent::MP;
8 8
9 NODE # returns this node identifier
10 $NODE # contains this node identifier 9 $NODE # contains this node's noderef
10 NODE # returns this node's noderef
11 NODE $port # returns the noderef of the port
11 12
13 $SELF # receiving/own port id in rcv callbacks
14
15 # initialise the node so it can send/receive messages
16 initialise_node; # -OR-
17 initialise_node "localhost:4040"; # -OR-
18 initialise_node "slave/", "localhost:4040"
19
20 # ports are message endpoints
21
22 # sending messages
12 snd $port, type => data...; 23 snd $port, type => data...;
24 snd $port, @msg;
25 snd @msg_with_first_element_being_a_port;
13 26
14 rcv $port, smartmatch => $cb->($port, @msg); 27 # creating/using ports, the simple way
28 my $simple_port = port { my @msg = @_; 0 };
15 29
16 # examples: 30 # creating/using ports, tagged message matching
31 my $port = port;
17 rcv $port2, ping => sub { snd $_[0], "pong"; 0 }; 32 rcv $port, ping => sub { snd $_[0], "pong"; 0 };
18 rcv $port1, pong => sub { warn "pong received\n" }; 33 rcv $port, pong => sub { warn "pong received\n"; 0 };
19 snd $port2, ping => $port1;
20 34
21 # more, smarter, matches (_any_ is exported by this module) 35 # create a port on another node
22 rcv $port, [child_died => $pid] => sub { ... 36 my $port = spawn $node, $initfunc, @initdata;
23 rcv $port, [_any_, _any_, 3] => sub { .. $_[2] is 3 37
38 # monitoring
39 mon $port, $cb->(@msg) # callback is invoked on death
40 mon $port, $otherport # kill otherport on abnormal death
41 mon $port, $otherport, @msg # send message on death
42
43=head1 CURRENT STATUS
44
45 AnyEvent::MP - stable API, should work
46 AnyEvent::MP::Intro - outdated
47 AnyEvent::MP::Kernel - WIP
48 AnyEvent::MP::Transport - mostly stable
49
50 stay tuned.
24 51
25=head1 DESCRIPTION 52=head1 DESCRIPTION
26 53
27This module (-family) implements a simple message passing framework. 54This module (-family) implements a simple message passing framework.
28 55
29Despite its simplicity, you can securely message other processes running 56Despite its simplicity, you can securely message other processes running
30on the same or other hosts. 57on the same or other hosts.
31 58
59For an introduction to this module family, see the L<AnyEvent::MP::Intro>
60manual page.
61
32At the moment, this module family is severly brokena nd underdocumented, 62At the moment, this module family is severly broken and underdocumented,
33so do not use. This was uploaded mainly to resreve the CPAN namespace - 63so do not use. This was uploaded mainly to reserve the CPAN namespace -
34stay tuned! 64stay tuned!
35 65
36=head1 CONCEPTS 66=head1 CONCEPTS
37 67
38=over 4 68=over 4
39 69
40=item port 70=item port
41 71
42A port is something you can send messages to with the C<snd> function, and 72A port is something you can send messages to (with the C<snd> function).
43you can register C<rcv> handlers with. All C<rcv> handlers will receive 73
74Ports allow you to register C<rcv> handlers that can match all or just
44messages they match, messages will not be queued. 75some messages. Messages will not be queued.
45 76
46=item port id - C<noderef#portname> 77=item port id - C<noderef#portname>
47 78
48A port id is always the noderef, a hash-mark (C<#>) as separator, followed 79A port ID is the concatenation of a noderef, a hash-mark (C<#>) as
49by a port name (a printable string of unspecified format). 80separator, and a port name (a printable string of unspecified format). An
81exception is the the node port, whose ID is identical to its node
82reference.
50 83
51=item node 84=item node
52 85
53A node is a single process containing at least one port - the node 86A node is a single process containing at least one port - the node port,
54port. You can send messages to node ports to let them create new ports, 87which provides nodes to manage each other remotely, and to create new
55among other things. 88ports.
56 89
57Initially, nodes are either private (single-process only) or hidden 90Nodes are either private (single-process only), slaves (connected to a
58(connected to a master node only). Only when they epxlicitly "become 91master node only) or public nodes (connectable from unrelated nodes).
59public" can you send them messages from unrelated other nodes.
60 92
61=item noderef - C<host:port,host:port...>, C<id@noderef>, C<id> 93=item noderef - C<host:port,host:port...>, C<id@noderef>, C<id>
62 94
63A noderef is a string that either uniquely identifies a given node (for 95A node reference is a string that either simply identifies the node (for
64private and hidden nodes), or contains a recipe on how to reach a given 96private and slave nodes), or contains a recipe on how to reach a given
65node (for public nodes). 97node (for public nodes).
66 98
99This recipe is simply a comma-separated list of C<address:port> pairs (for
100TCP/IP, other protocols might look different).
101
102Node references come in two flavours: resolved (containing only numerical
103addresses) or unresolved (where hostnames are used instead of addresses).
104
105Before using an unresolved node reference in a message you first have to
106resolve it.
107
67=back 108=back
68 109
69=head1 VARIABLES/FUNCTIONS 110=head1 VARIABLES/FUNCTIONS
70 111
71=over 4 112=over 4
72 113
73=cut 114=cut
74 115
75package AnyEvent::MP; 116package AnyEvent::MP;
76 117
77use AnyEvent::MP::Util ();
78use AnyEvent::MP::Node; 118use AnyEvent::MP::Kernel;
79use AnyEvent::MP::Transport;
80 119
81use utf8;
82use common::sense; 120use common::sense;
83 121
84use Carp (); 122use Carp ();
85 123
86use AE (); 124use AE ();
87 125
88use base "Exporter"; 126use base "Exporter";
89 127
90our $VERSION = '0.01'; 128our $VERSION = $AnyEvent::MP::Kernel::VERSION;
91our @EXPORT = qw(NODE $NODE $PORT snd rcv _any_);
92 129
93our $DEFAULT_SECRET; 130our @EXPORT = qw(
94our $DEFAULT_PORT = "4040"; 131 NODE $NODE *SELF node_of _any_
132 resolve_node initialise_node
133 snd rcv mon kil reg psub spawn
134 port
135);
95 136
96our $CONNECT_INTERVAL = 5; # new connect every 5s, at least 137our $SELF;
97our $CONNECT_TIMEOUT = 30; # includes handshake
98 138
99sub default_secret { 139sub _self_die() {
100 unless (defined $DEFAULT_SECRET) { 140 my $msg = $@;
101 if (open my $fh, "<$ENV{HOME}/.aemp-secret") { 141 $msg =~ s/\n+$// unless ref $msg;
102 sysread $fh, $DEFAULT_SECRET, -s $fh; 142 kil $SELF, die => $msg;
103 } else {
104 $DEFAULT_SECRET = AnyEvent::MP::Util::nonce 32;
105 }
106 }
107
108 $DEFAULT_SECRET
109} 143}
110 144
111=item NODE / $NODE 145=item $thisnode = NODE / $NODE
112 146
113The C<NODE ()> function and the C<$NODE> variable contain the noderef of 147The C<NODE> function returns, and the C<$NODE> variable contains the
114the local node. The value is initialised by a call to C<become_public> or 148noderef of the local node. The value is initialised by a call to
115C<become_slave>, after which all local port identifiers become invalid. 149C<initialise_node>.
116 150
117=cut 151=item $noderef = node_of $port
118 152
119our $UNIQ = sprintf "%x.%x", $$, time; # per-process/node unique cookie 153Extracts and returns the noderef from a port ID or a noderef.
120our $ID = "a0";
121our $PUBLIC = 0;
122our $NODE;
123our $PORT;
124 154
125our %NODE; # node id to transport mapping, or "undef", for local node 155=item initialise_node $noderef, $seednode, $seednode...
126our %PORT; # local ports
127our %LISTENER; # local transports
128 156
129sub NODE() { $NODE } 157=item initialise_node "slave/", $master, $master...
130 158
131{ 159Before a node can talk to other nodes on the network it has to initialise
132 use POSIX (); 160itself - the minimum a node needs to know is it's own name, and optionally
133 my $nodename = (POSIX::uname)[1]; 161it should know the noderefs of some other nodes in the network.
134 $NODE = "$$\@$nodename";
135}
136 162
137sub _ANY_() { 1 } 163This function initialises a node - it must be called exactly once (or
138sub _any_() { \&_ANY_ } 164never) before calling other AnyEvent::MP functions.
139 165
140sub add_node { 166All arguments (optionally except for the first) are noderefs, which can be
141 my ($noderef) = @_; 167either resolved or unresolved.
142 168
143 return $NODE{$noderef} 169The first argument will be looked up in the configuration database first
144 if exists $NODE{$noderef}; 170(if it is C<undef> then the current nodename will be used instead) to find
171the relevant configuration profile (see L<aemp>). If none is found then
172the default configuration is used. The configuration supplies additional
173seed/master nodes and can override the actual noderef.
145 174
146 for (split /,/, $noderef) { 175There are two types of networked nodes, public nodes and slave nodes:
147 return $NODE{$noderef} = $NODE{$_}
148 if exists $NODE{$_};
149 }
150 176
151 # for indirect sends, use a different class 177=over 4
152 my $node = new AnyEvent::MP::Node::Direct $noderef;
153 178
154 $NODE{$_} = $node 179=item public nodes
155 for $noderef, split /,/, $noderef;
156 180
157 $node 181For public nodes, C<$noderef> (supplied either directly to
158} 182C<initialise_node> or indirectly via a profile or the nodename) must be a
183noderef (possibly unresolved, in which case it will be resolved).
159 184
185After resolving, the node will bind itself on all endpoints and try to
186connect to all additional C<$seednodes> that are specified. Seednodes are
187optional and can be used to quickly bootstrap the node into an existing
188network.
189
190=item slave nodes
191
192When the C<$noderef> (either as given or overriden by the config file)
193is the special string C<slave/>, then the node will become a slave
194node. Slave nodes cannot be contacted from outside and will route most of
195their traffic to the master node that they attach to.
196
197At least one additional noderef is required (either by specifying it
198directly or because it is part of the configuration profile): The node
199will try to connect to all of them and will become a slave attached to the
200first node it can successfully connect to.
201
202=back
203
204This function will block until all nodes have been resolved and, for slave
205nodes, until it has successfully established a connection to a master
206server.
207
208Example: become a public node listening on the guessed noderef, or the one
209specified via C<aemp> for the current node. This should be the most common
210form of invocation for "daemon"-type nodes.
211
212 initialise_node;
213
214Example: become a slave node to any of the the seednodes specified via
215C<aemp>. This form is often used for commandline clients.
216
217 initialise_node "slave/";
218
219Example: become a slave node to any of the specified master servers. This
220form is also often used for commandline clients.
221
222 initialise_node "slave/", "master1", "192.168.13.17", "mp.example.net";
223
224Example: become a public node, and try to contact some well-known master
225servers to become part of the network.
226
227 initialise_node undef, "master1", "master2";
228
229Example: become a public node listening on port C<4041>.
230
231 initialise_node 4041;
232
233Example: become a public node, only visible on localhost port 4044.
234
235 initialise_node "localhost:4044";
236
237=item $cv = resolve_node $noderef
238
239Takes an unresolved node reference that may contain hostnames and
240abbreviated IDs, resolves all of them and returns a resolved node
241reference.
242
243In addition to C<address:port> pairs allowed in resolved noderefs, the
244following forms are supported:
245
246=over 4
247
248=item the empty string
249
250An empty-string component gets resolved as if the default port (4040) was
251specified.
252
253=item naked port numbers (e.g. C<1234>)
254
255These are resolved by prepending the local nodename and a colon, to be
256further resolved.
257
258=item hostnames (e.g. C<localhost:1234>, C<localhost>)
259
260These are resolved by using AnyEvent::DNS to resolve them, optionally
261looking up SRV records for the C<aemp=4040> port, if no port was
262specified.
263
264=back
265
266=item $SELF
267
268Contains the current port id while executing C<rcv> callbacks or C<psub>
269blocks.
270
271=item SELF, %SELF, @SELF...
272
273Due to some quirks in how perl exports variables, it is impossible to
274just export C<$SELF>, all the symbols called C<SELF> are exported by this
275module, but only C<$SELF> is currently used.
276
160=item snd $portid, type => @data 277=item snd $port, type => @data
161 278
162=item snd $portid, @msg 279=item snd $port, @msg
163 280
164Send the given message to the given port ID, which can identify either a 281Send the given message to the given port ID, which can identify either
165local or a remote port. 282a local or a remote port, and must be a port ID.
166 283
167While the message can be about anything, it is highly recommended to use 284While the message can be about anything, it is highly recommended to use a
168a constant string as first element. 285string as first element (a port ID, or some word that indicates a request
286type etc.).
169 287
170The message data effectively becomes read-only after a call to this 288The message data effectively becomes read-only after a call to this
171function: modifying any argument is not allowed and can cause many 289function: modifying any argument is not allowed and can cause many
172problems. 290problems.
173 291
175JSON is used, then only strings, numbers and arrays and hashes consisting 293JSON is used, then only strings, numbers and arrays and hashes consisting
176of those are allowed (no objects). When Storable is used, then anything 294of those are allowed (no objects). When Storable is used, then anything
177that Storable can serialise and deserialise is allowed, and for the local 295that Storable can serialise and deserialise is allowed, and for the local
178node, anything can be passed. 296node, anything can be passed.
179 297
180=cut 298=item $local_port = port
181 299
182sub snd(@) { 300Create a new local port object and returns its port ID. Initially it has
183 my ($noderef, $port) = split /#/, shift, 2; 301no callbacks set and will throw an error when it receives messages.
184 302
185 add_node $noderef 303=item $local_port = port { my @msg = @_ }
186 unless exists $NODE{$noderef};
187 304
188 $NODE{$noderef}->send (["$port", [@_]]); 305Creates a new local port, and returns its ID. Semantically the same as
306creating a port and calling C<rcv $port, $callback> on it.
307
308The block will be called for every message received on the port, with the
309global variable C<$SELF> set to the port ID. Runtime errors will cause the
310port to be C<kil>ed. The message will be passed as-is, no extra argument
311(i.e. no port ID) will be passed to the callback.
312
313If you want to stop/destroy the port, simply C<kil> it:
314
315 my $port = port {
316 my @msg = @_;
317 ...
318 kil $SELF;
319 };
320
321=cut
322
323sub rcv($@);
324
325sub _kilme {
326 die "received message on port without callback";
189} 327}
190 328
329sub port(;&) {
330 my $id = "$UNIQ." . $ID++;
331 my $port = "$NODE#$id";
332
333 rcv $port, shift || \&_kilme;
334
335 $port
336}
337
191=item rcv $portid, type => $callback->(@msg) 338=item rcv $local_port, $callback->(@msg)
192 339
193=item rcv $portid, $smartmatch => $callback->(@msg) 340Replaces the default callback on the specified port. There is no way to
341remove the default callback: use C<sub { }> to disable it, or better
342C<kil> the port when it is no longer needed.
194 343
195=item rcv $portid, [$smartmatch...] => $callback->(@msg) 344The global C<$SELF> (exported by this module) contains C<$port> while
345executing the callback. Runtime errors during callback execution will
346result in the port being C<kil>ed.
196 347
197Register a callback on the port identified by C<$portid>, which I<must> be 348The default callback received all messages not matched by a more specific
198a local port. 349C<tag> match.
199 350
200The callback has to return a true value when its work is done, after 351=item rcv $local_port, tag => $callback->(@msg_without_tag), ...
201which is will be removed, or a false value in which case it will stay
202registered.
203 352
204If the match is an array reference, then it will be matched against the 353Register (or replace) callbacks to be called on messages starting with the
205first elements of the message, otherwise only the first element is being 354given tag on the given port (and return the port), or unregister it (when
206matched. 355C<$callback> is C<$undef> or missing). There can only be one callback
356registered for each tag.
207 357
208Any element in the match that is specified as C<_any_> (a function 358The original message will be passed to the callback, after the first
209exported by this module) matches any single element of the message. 359element (the tag) has been removed. The callback will use the same
360environment as the default callback (see above).
210 361
211While not required, it is highly recommended that the first matching 362Example: create a port and bind receivers on it in one go.
212element is a string identifying the message. The one-string-only match is 363
213also the most efficient match (by far). 364 my $port = rcv port,
365 msg1 => sub { ... },
366 msg2 => sub { ... },
367 ;
368
369Example: create a port, bind receivers and send it in a message elsewhere
370in one go:
371
372 snd $otherport, reply =>
373 rcv port,
374 msg1 => sub { ... },
375 ...
376 ;
377
378Example: temporarily register a rcv callback for a tag matching some port
379(e.g. for a rpc reply) and unregister it after a message was received.
380
381 rcv $port, $otherport => sub {
382 my @reply = @_;
383
384 rcv $SELF, $otherport;
385 };
214 386
215=cut 387=cut
216 388
217sub rcv($@) { 389sub rcv($@) {
218 my ($port, $match, $cb) = @_; 390 my $port = shift;
219
220 my $port = $PORT{$port}
221 or do {
222 my ($noderef, $lport) = split /#/, $port; 391 my ($noderef, $portid) = split /#/, $port, 2;
223 "AnyEvent::MP::Node::Self" eq ref $NODE{$noderef} 392
393 ($NODE{$noderef} || add_node $noderef) == $NODE{""}
224 or Carp::croak "$port: can only rcv on local ports"; 394 or Carp::croak "$port: rcv can only be called on local ports, caught";
225 395
226 $PORT{$lport} 396 while (@_) {
227 or Carp::croak "$port: port does not exist"; 397 if (ref $_[0]) {
228 398 if (my $self = $PORT_DATA{$portid}) {
229 $PORT{$port} = $PORT{$lport} # also return 399 "AnyEvent::MP::Port" eq ref $self
230 }; 400 or Carp::croak "$port: rcv can only be called on message matching ports, caught";
231 401
232 if (!ref $match) { 402 $self->[2] = shift;
233 push @{ $port->{rc0}{$match} }, [$cb];
234 } elsif (("ARRAY" eq ref $match && !ref $match->[0])) {
235 my ($type, @match) = @$match;
236 @match
237 ? push @{ $port->{rcv}{$match->[0]} }, [$cb, \@match]
238 : push @{ $port->{rc0}{$match->[0]} }, [$cb];
239 } else { 403 } else {
240 push @{ $port->{any} }, [$cb, $match]; 404 my $cb = shift;
241 } 405 $PORT{$portid} = sub {
242} 406 local $SELF = $port;
243 407 eval { &$cb }; _self_die if $@;
244sub _inject {
245 my ($port, $msg) = @{+shift};
246
247 $port = $PORT{$port}
248 or return;
249
250 @_ = @$msg;
251
252 for (@{ $port->{rc0}{$msg->[0]} }) {
253 $_ && &{$_->[0]}
254 && undef $_;
255 }
256
257 for (@{ $port->{rcv}{$msg->[0]} }) {
258 $_ && [@_[1..$#{$_->[1]}]] ~~ $_->[1]
259 && &{$_->[0]}
260 && undef $_;
261 }
262
263 for (@{ $port->{any} }) {
264 $_ && [@_[0..$#{$_->[1]}]] ~~ $_->[1]
265 && &{$_->[0]}
266 && undef $_;
267 }
268}
269
270sub normalise_noderef($) {
271 my ($noderef) = @_;
272
273 my $cv = AE::cv;
274 my @res;
275
276 $cv->begin (sub {
277 my %seen;
278 my @refs;
279 for (sort { $a->[0] <=> $b->[0] } @res) {
280 push @refs, $_->[1] unless $seen{$_->[1]}++
281 }
282 shift->send (join ",", @refs);
283 });
284
285 $noderef = $DEFAULT_PORT unless length $noderef;
286
287 my $idx;
288 for my $t (split /,/, $noderef) {
289 my $pri = ++$idx;
290
291 #TODO: this should be outside normalise_noderef and in become_public
292 if ($t =~ /^\d*$/) {
293 my $nodename = (POSIX::uname)[1];
294
295 $cv->begin;
296 AnyEvent::Socket::resolve_sockaddr $nodename, $t || "aemp=$DEFAULT_PORT", "tcp", 0, undef, sub {
297 for (@_) {
298 my ($service, $host) = AnyEvent::Socket::unpack_sockaddr $_->[3];
299 push @res, [
300 $pri += 1e-5,
301 AnyEvent::Socket::format_hostport AnyEvent::Socket::format_address $host, $service
302 ];
303 } 408 };
304 $cv->end; 409 }
410 } elsif (defined $_[0]) {
411 my $self = $PORT_DATA{$portid} ||= do {
412 my $self = bless [$PORT{$port} || sub { }, { }, $port], "AnyEvent::MP::Port";
413
414 $PORT{$portid} = sub {
415 local $SELF = $port;
416
417 if (my $cb = $self->[1]{$_[0]}) {
418 shift;
419 eval { &$cb }; _self_die if $@;
420 } else {
421 &{ $self->[0] };
422 }
423 };
424
425 $self
305 }; 426 };
306 427
307# my (undef, undef, undef, undef, @ipv4) = gethostbyname $nodename; 428 "AnyEvent::MP::Port" eq ref $self
308# 429 or Carp::croak "$port: rcv can only be called on message matching ports, caught";
309# for (@ipv4) { 430
310# push @res, [ 431 my ($tag, $cb) = splice @_, 0, 2;
311# $pri, 432
312# AnyEvent::Socket::format_hostport AnyEvent::Socket::format_address $_, $t || $DEFAULT_PORT, 433 if (defined $cb) {
313# ]; 434 $self->[1]{$tag} = $cb;
314# }
315 } else { 435 } else {
316 my ($host, $port) = AnyEvent::Socket::parse_hostport $t, "aemp=$DEFAULT_PORT" 436 delete $self->[1]{$tag};
317 or Carp::croak "$t: unparsable transport descriptor";
318
319 $cv->begin;
320 AnyEvent::Socket::resolve_sockaddr $host, $port, "tcp", 0, undef, sub {
321 for (@_) {
322 my ($service, $host) = AnyEvent::Socket::unpack_sockaddr $_->[3];
323 push @res, [
324 $pri += 1e-5,
325 AnyEvent::Socket::format_hostport AnyEvent::Socket::format_address $host, $service
326 ];
327 }
328 $cv->end;
329 } 437 }
330 } 438 }
331 } 439 }
332 440
333 $cv->end; 441 $port
334
335 $cv
336} 442}
337 443
338sub become_public { 444=item $closure = psub { BLOCK }
339 return if $PUBLIC;
340 445
341 my $noderef = join ",", ref $_[0] ? @{+shift} : shift; 446Remembers C<$SELF> and creates a closure out of the BLOCK. When the
342 my @args = @_; 447closure is executed, sets up the environment in the same way as in C<rcv>
448callbacks, i.e. runtime errors will cause the port to get C<kil>ed.
343 449
344 $NODE = (normalise_noderef $noderef)->recv; 450This is useful when you register callbacks from C<rcv> callbacks:
345 451
346 for my $t (split /,/, $NODE) { 452 rcv delayed_reply => sub {
347 $NODE{$t} = $NODE{""}; 453 my ($delay, @reply) = @_;
348 454 my $timer = AE::timer $delay, 0, psub {
349 my ($host, $port) = AnyEvent::Socket::parse_hostport $t; 455 snd @reply, $SELF;
350
351 $LISTENER{$t} = AnyEvent::MP::Transport::mp_server $host, $port,
352 @args,
353 on_error => sub {
354 die "on_error<@_>\n";#d#
355 },
356 on_connect => sub {
357 my ($tp) = @_;
358
359 $NODE{$tp->{remote_id}} = $_[0];
360 },
361 sub {
362 my ($tp) = @_;
363
364 $NODE{"$tp->{peerhost}:$tp->{peerport}"} = $tp;
365 },
366 ; 456 };
457 };
458
459=cut
460
461sub psub(&) {
462 my $cb = shift;
463
464 my $port = $SELF
465 or Carp::croak "psub can only be called from within rcv or psub callbacks, not";
466
467 sub {
468 local $SELF = $port;
469
470 if (wantarray) {
471 my @res = eval { &$cb };
472 _self_die if $@;
473 @res
474 } else {
475 my $res = eval { &$cb };
476 _self_die if $@;
477 $res
478 }
367 } 479 }
480}
368 481
369 $PUBLIC = 1; 482=item $guard = mon $port, $cb->(@reason)
483
484=item $guard = mon $port, $rcvport
485
486=item $guard = mon $port
487
488=item $guard = mon $port, $rcvport, @msg
489
490Monitor the given port and do something when the port is killed or
491messages to it were lost, and optionally return a guard that can be used
492to stop monitoring again.
493
494C<mon> effectively guarantees that, in the absence of hardware failures,
495that after starting the monitor, either all messages sent to the port
496will arrive, or the monitoring action will be invoked after possible
497message loss has been detected. No messages will be lost "in between"
498(after the first lost message no further messages will be received by the
499port). After the monitoring action was invoked, further messages might get
500delivered again.
501
502In the first form (callback), the callback is simply called with any
503number of C<@reason> elements (no @reason means that the port was deleted
504"normally"). Note also that I<< the callback B<must> never die >>, so use
505C<eval> if unsure.
506
507In the second form (another port given), the other port (C<$rcvport>)
508will be C<kil>'ed with C<@reason>, iff a @reason was specified, i.e. on
509"normal" kils nothing happens, while under all other conditions, the other
510port is killed with the same reason.
511
512The third form (kill self) is the same as the second form, except that
513C<$rvport> defaults to C<$SELF>.
514
515In the last form (message), a message of the form C<@msg, @reason> will be
516C<snd>.
517
518As a rule of thumb, monitoring requests should always monitor a port from
519a local port (or callback). The reason is that kill messages might get
520lost, just like any other message. Another less obvious reason is that
521even monitoring requests can get lost (for exmaple, when the connection
522to the other node goes down permanently). When monitoring a port locally
523these problems do not exist.
524
525Example: call a given callback when C<$port> is killed.
526
527 mon $port, sub { warn "port died because of <@_>\n" };
528
529Example: kill ourselves when C<$port> is killed abnormally.
530
531 mon $port;
532
533Example: send us a restart message when another C<$port> is killed.
534
535 mon $port, $self => "restart";
536
537=cut
538
539sub mon {
540 my ($noderef, $port) = split /#/, shift, 2;
541
542 my $node = $NODE{$noderef} || add_node $noderef;
543
544 my $cb = @_ ? shift : $SELF || Carp::croak 'mon: called with one argument only, but $SELF not set,';
545
546 unless (ref $cb) {
547 if (@_) {
548 # send a kill info message
549 my (@msg) = ($cb, @_);
550 $cb = sub { snd @msg, @_ };
551 } else {
552 # simply kill other port
553 my $port = $cb;
554 $cb = sub { kil $port, @_ if @_ };
555 }
556 }
557
558 $node->monitor ($port, $cb);
559
560 defined wantarray
561 and AnyEvent::Util::guard { $node->unmonitor ($port, $cb) }
562}
563
564=item $guard = mon_guard $port, $ref, $ref...
565
566Monitors the given C<$port> and keeps the passed references. When the port
567is killed, the references will be freed.
568
569Optionally returns a guard that will stop the monitoring.
570
571This function is useful when you create e.g. timers or other watchers and
572want to free them when the port gets killed:
573
574 $port->rcv (start => sub {
575 my $timer; $timer = mon_guard $port, AE::timer 1, 1, sub {
576 undef $timer if 0.9 < rand;
577 });
578 });
579
580=cut
581
582sub mon_guard {
583 my ($port, @refs) = @_;
584
585 #TODO: mon-less form?
586
587 mon $port, sub { 0 && @refs }
588}
589
590=item kil $port[, @reason]
591
592Kill the specified port with the given C<@reason>.
593
594If no C<@reason> is specified, then the port is killed "normally" (linked
595ports will not be kileld, or even notified).
596
597Otherwise, linked ports get killed with the same reason (second form of
598C<mon>, see below).
599
600Runtime errors while evaluating C<rcv> callbacks or inside C<psub> blocks
601will be reported as reason C<< die => $@ >>.
602
603Transport/communication errors are reported as C<< transport_error =>
604$message >>.
605
606=cut
607
608=item $port = spawn $node, $initfunc[, @initdata]
609
610Creates a port on the node C<$node> (which can also be a port ID, in which
611case it's the node where that port resides).
612
613The port ID of the newly created port is return immediately, and it is
614permissible to immediately start sending messages or monitor the port.
615
616After the port has been created, the init function is
617called. This function must be a fully-qualified function name
618(e.g. C<MyApp::Chat::Server::init>). To specify a function in the main
619program, use C<::name>.
620
621If the function doesn't exist, then the node tries to C<require>
622the package, then the package above the package and so on (e.g.
623C<MyApp::Chat::Server>, C<MyApp::Chat>, C<MyApp>) until the function
624exists or it runs out of package names.
625
626The init function is then called with the newly-created port as context
627object (C<$SELF>) and the C<@initdata> values as arguments.
628
629A common idiom is to pass your own port, monitor the spawned port, and
630in the init function, monitor the original port. This two-way monitoring
631ensures that both ports get cleaned up when there is a problem.
632
633Example: spawn a chat server port on C<$othernode>.
634
635 # this node, executed from within a port context:
636 my $server = spawn $othernode, "MyApp::Chat::Server::connect", $SELF;
637 mon $server;
638
639 # init function on C<$othernode>
640 sub connect {
641 my ($srcport) = @_;
642
643 mon $srcport;
644
645 rcv $SELF, sub {
646 ...
647 };
648 }
649
650=cut
651
652sub _spawn {
653 my $port = shift;
654 my $init = shift;
655
656 local $SELF = "$NODE#$port";
657 eval {
658 &{ load_func $init }
659 };
660 _self_die if $@;
661}
662
663sub spawn(@) {
664 my ($noderef, undef) = split /#/, shift, 2;
665
666 my $id = "$RUNIQ." . $ID++;
667
668 $_[0] =~ /::/
669 or Carp::croak "spawn init function must be a fully-qualified name, caught";
670
671 ($NODE{$noderef} || add_node $noderef)
672 ->send (["", "AnyEvent::MP::_spawn" => $id, @_]);
673
674 "$noderef#$id"
370} 675}
371 676
372=back 677=back
373 678
374=head1 NODE MESSAGES 679=head1 NODE MESSAGES
376Nodes understand the following messages sent to them. Many of them take 681Nodes understand the following messages sent to them. Many of them take
377arguments called C<@reply>, which will simply be used to compose a reply 682arguments called C<@reply>, which will simply be used to compose a reply
378message - C<$reply[0]> is the port to reply to, C<$reply[1]> the type and 683message - C<$reply[0]> is the port to reply to, C<$reply[1]> the type and
379the remaining arguments are simply the message data. 684the remaining arguments are simply the message data.
380 685
686While other messages exist, they are not public and subject to change.
687
381=over 4 688=over 4
382 689
383=cut 690=cut
384 691
385############################################################################# 692=item lookup => $name, @reply
386# self node code
387 693
388sub _new_port($) { 694Replies with the port ID of the specified well-known port, or C<undef>.
389 my ($name) = @_;
390
391 my ($noderef, $portname) = split /#/, $name;
392
393 $PORT{$name} =
394 $PORT{$portname} = {
395 names => [$name, $portname],
396 };
397}
398
399$NODE{""} = new AnyEvent::MP::Node::Self noderef => $NODE;
400_new_port "";
401 695
402=item devnull => ... 696=item devnull => ...
403 697
404Generic data sink/CPU heat conversion. 698Generic data sink/CPU heat conversion.
405 699
406=cut
407
408rcv "", devnull => sub { () };
409
410=item relay => $port, @msg 700=item relay => $port, @msg
411 701
412Simply forwards the message to the given port. 702Simply forwards the message to the given port.
413
414=cut
415
416rcv "", relay => sub { \&snd; () };
417 703
418=item eval => $string[ @reply] 704=item eval => $string[ @reply]
419 705
420Evaluates the given string. If C<@reply> is given, then a message of the 706Evaluates the given string. If C<@reply> is given, then a message of the
421form C<@reply, $@, @evalres> is sent. 707form C<@reply, $@, @evalres> is sent.
422 708
423Example: crash another node. 709Example: crash another node.
424 710
425 snd $othernode, eval => "exit"; 711 snd $othernode, eval => "exit";
426 712
427=cut
428
429rcv "", eval => sub {
430 my (undef, $string, @reply) = @_;
431 my @res = eval $string;
432 snd @reply, "$@", @res if @reply;
433 ()
434};
435
436=item time => @reply 713=item time => @reply
437 714
438Replies the the current node time to C<@reply>. 715Replies the the current node time to C<@reply>.
439 716
440Example: tell the current node to send the current time to C<$myport> in a 717Example: tell the current node to send the current time to C<$myport> in a
441C<timereply> message. 718C<timereply> message.
442 719
443 snd $NODE, time => $myport, timereply => 1, 2; 720 snd $NODE, time => $myport, timereply => 1, 2;
444 # => snd $myport, timereply => 1, 2, <time> 721 # => snd $myport, timereply => 1, 2, <time>
445 722
446=cut 723=back
447 724
448rcv "", time => sub { shift; snd @_, AE::time; () }; 725=head1 AnyEvent::MP vs. Distributed Erlang
726
727AnyEvent::MP got lots of its ideas from distributed Erlang (Erlang node
728== aemp node, Erlang process == aemp port), so many of the documents and
729programming techniques employed by Erlang apply to AnyEvent::MP. Here is a
730sample:
731
732 http://www.Erlang.se/doc/programming_rules.shtml
733 http://Erlang.org/doc/getting_started/part_frame.html # chapters 3 and 4
734 http://Erlang.org/download/Erlang-book-part1.pdf # chapters 5 and 6
735 http://Erlang.org/download/armstrong_thesis_2003.pdf # chapters 4 and 5
736
737Despite the similarities, there are also some important differences:
738
739=over 4
740
741=item * Node references contain the recipe on how to contact them.
742
743Erlang relies on special naming and DNS to work everywhere in the
744same way. AEMP relies on each node knowing it's own address(es), with
745convenience functionality.
746
747This means that AEMP requires a less tightly controlled environment at the
748cost of longer node references and a slightly higher management overhead.
749
750=item * Erlang has a "remote ports are like local ports" philosophy, AEMP
751uses "local ports are like remote ports".
752
753The failure modes for local ports are quite different (runtime errors
754only) then for remote ports - when a local port dies, you I<know> it dies,
755when a connection to another node dies, you know nothing about the other
756port.
757
758Erlang pretends remote ports are as reliable as local ports, even when
759they are not.
760
761AEMP encourages a "treat remote ports differently" philosophy, with local
762ports being the special case/exception, where transport errors cannot
763occur.
764
765=item * Erlang uses processes and a mailbox, AEMP does not queue.
766
767Erlang uses processes that selectively receive messages, and therefore
768needs a queue. AEMP is event based, queuing messages would serve no
769useful purpose. For the same reason the pattern-matching abilities of
770AnyEvent::MP are more limited, as there is little need to be able to
771filter messages without dequeing them.
772
773(But see L<Coro::MP> for a more Erlang-like process model on top of AEMP).
774
775=item * Erlang sends are synchronous, AEMP sends are asynchronous.
776
777Sending messages in Erlang is synchronous and blocks the process (and
778so does not need a queue that can overflow). AEMP sends are immediate,
779connection establishment is handled in the background.
780
781=item * Erlang suffers from silent message loss, AEMP does not.
782
783Erlang makes few guarantees on messages delivery - messages can get lost
784without any of the processes realising it (i.e. you send messages a, b,
785and c, and the other side only receives messages a and c).
786
787AEMP guarantees correct ordering, and the guarantee that there are no
788holes in the message sequence.
789
790=item * In Erlang, processes can be declared dead and later be found to be
791alive.
792
793In Erlang it can happen that a monitored process is declared dead and
794linked processes get killed, but later it turns out that the process is
795still alive - and can receive messages.
796
797In AEMP, when port monitoring detects a port as dead, then that port will
798eventually be killed - it cannot happen that a node detects a port as dead
799and then later sends messages to it, finding it is still alive.
800
801=item * Erlang can send messages to the wrong port, AEMP does not.
802
803In Erlang it is quite likely that a node that restarts reuses a process ID
804known to other nodes for a completely different process, causing messages
805destined for that process to end up in an unrelated process.
806
807AEMP never reuses port IDs, so old messages or old port IDs floating
808around in the network will not be sent to an unrelated port.
809
810=item * Erlang uses unprotected connections, AEMP uses secure
811authentication and can use TLS.
812
813AEMP can use a proven protocol - SSL/TLS - to protect connections and
814securely authenticate nodes.
815
816=item * The AEMP protocol is optimised for both text-based and binary
817communications.
818
819The AEMP protocol, unlike the Erlang protocol, supports both
820language-independent text-only protocols (good for debugging) and binary,
821language-specific serialisers (e.g. Storable).
822
823It has also been carefully designed to be implementable in other languages
824with a minimum of work while gracefully degrading fucntionality to make the
825protocol simple.
826
827=item * AEMP has more flexible monitoring options than Erlang.
828
829In Erlang, you can chose to receive I<all> exit signals as messages
830or I<none>, there is no in-between, so monitoring single processes is
831difficult to implement. Monitoring in AEMP is more flexible than in
832Erlang, as one can choose between automatic kill, exit message or callback
833on a per-process basis.
834
835=item * Erlang tries to hide remote/local connections, AEMP does not.
836
837Monitoring in Erlang is not an indicator of process death/crashes,
838as linking is (except linking is unreliable in Erlang).
839
840In AEMP, you don't "look up" registered port names or send to named ports
841that might or might not be persistent. Instead, you normally spawn a port
842on the remote node. The init function monitors the you, and you monitor
843the remote port. Since both monitors are local to the node, they are much
844more reliable.
845
846This also saves round-trips and avoids sending messages to the wrong port
847(hard to do in Erlang).
848
849=back
850
851=head1 RATIONALE
852
853=over 4
854
855=item Why strings for ports and noderefs, why not objects?
856
857We considered "objects", but found that the actual number of methods
858thatc an be called are very low. Since port IDs and noderefs travel over
859the network frequently, the serialising/deserialising would add lots of
860overhead, as well as having to keep a proxy object.
861
862Strings can easily be printed, easily serialised etc. and need no special
863procedures to be "valid".
864
865And a a miniport consists of a single closure stored in a global hash - it
866can't become much cheaper.
867
868=item Why favour JSON, why not real serialising format such as Storable?
869
870In fact, any AnyEvent::MP node will happily accept Storable as framing
871format, but currently there is no way to make a node use Storable by
872default.
873
874The default framing protocol is JSON because a) JSON::XS is many times
875faster for small messages and b) most importantly, after years of
876experience we found that object serialisation is causing more problems
877than it gains: Just like function calls, objects simply do not travel
878easily over the network, mostly because they will always be a copy, so you
879always have to re-think your design.
880
881Keeping your messages simple, concentrating on data structures rather than
882objects, will keep your messages clean, tidy and efficient.
449 883
450=back 884=back
451 885
452=head1 SEE ALSO 886=head1 SEE ALSO
453 887

Diff Legend

Removed lines
+ Added lines
< Changed lines
> Changed lines