1 | =head1 NAME |
1 | =head1 NAME |
2 | |
2 | |
3 | AnyEvent::MP - multi-processing/message-passing framework |
3 | AnyEvent::MP - erlang-style multi-processing/message-passing framework |
4 | |
4 | |
5 | =head1 SYNOPSIS |
5 | =head1 SYNOPSIS |
6 | |
6 | |
7 | use AnyEvent::MP; |
7 | use AnyEvent::MP; |
8 | |
8 | |
9 | $NODE # contains this node's noderef |
9 | $NODE # contains this node's node ID |
10 | NODE # returns this node's noderef |
10 | NODE # returns this node's node ID |
11 | NODE $port # returns the noderef of the port |
|
|
12 | |
11 | |
13 | $SELF # receiving/own port id in rcv callbacks |
12 | $SELF # receiving/own port id in rcv callbacks |
14 | |
13 | |
15 | # initialise the node so it can send/receive messages |
14 | # initialise the node so it can send/receive messages |
16 | initialise_node; # -OR- |
15 | configure; |
17 | initialise_node "localhost:4040"; # -OR- |
|
|
18 | initialise_node "slave/", "localhost:4040" |
|
|
19 | |
16 | |
20 | # ports are message endpoints |
17 | # ports are message destinations |
21 | |
18 | |
22 | # sending messages |
19 | # sending messages |
23 | snd $port, type => data...; |
20 | snd $port, type => data...; |
24 | snd $port, @msg; |
21 | snd $port, @msg; |
25 | snd @msg_with_first_element_being_a_port; |
22 | snd @msg_with_first_element_being_a_port; |
26 | |
23 | |
27 | # creating/using ports, the simple way |
24 | # creating/using ports, the simple way |
28 | my $somple_port = port { my @msg = @_; 0 }; |
25 | my $simple_port = port { my @msg = @_ }; |
29 | |
26 | |
30 | # creating/using ports, tagged message matching |
27 | # creating/using ports, tagged message matching |
31 | my $port = port; |
28 | my $port = port; |
32 | rcv $port, ping => sub { snd $_[0], "pong"; 0 }; |
29 | rcv $port, ping => sub { snd $_[0], "pong" }; |
33 | rcv $port, pong => sub { warn "pong received\n"; 0 }; |
30 | rcv $port, pong => sub { warn "pong received\n" }; |
34 | |
31 | |
35 | # create a port on another node |
32 | # create a port on another node |
36 | my $port = spawn $node, $initfunc, @initdata; |
33 | my $port = spawn $node, $initfunc, @initdata; |
37 | |
34 | |
|
|
35 | # destroy a prot again |
|
|
36 | kil $port; # "normal" kill |
|
|
37 | kil $port, my_error => "everything is broken"; # error kill |
|
|
38 | |
38 | # monitoring |
39 | # monitoring |
39 | mon $port, $cb->(@msg) # callback is invoked on death |
40 | mon $localport, $cb->(@msg) # callback is invoked on death |
40 | mon $port, $otherport # kill otherport on abnormal death |
41 | mon $localport, $otherport # kill otherport on abnormal death |
41 | mon $port, $otherport, @msg # send message on death |
42 | mon $localport, $otherport, @msg # send message on death |
|
|
43 | |
|
|
44 | # temporarily execute code in port context |
|
|
45 | peval $port, sub { die "kill the port!" }; |
|
|
46 | |
|
|
47 | # execute callbacks in $SELF port context |
|
|
48 | my $timer = AE::timer 1, 0, psub { |
|
|
49 | die "kill the port, delayed"; |
|
|
50 | }; |
42 | |
51 | |
43 | =head1 CURRENT STATUS |
52 | =head1 CURRENT STATUS |
44 | |
53 | |
|
|
54 | bin/aemp - stable. |
45 | AnyEvent::MP - stable API, should work |
55 | AnyEvent::MP - stable API, should work. |
46 | AnyEvent::MP::Intro - outdated |
56 | AnyEvent::MP::Intro - explains most concepts. |
47 | AnyEvent::MP::Kernel - WIP |
|
|
48 | AnyEvent::MP::Transport - mostly stable |
57 | AnyEvent::MP::Kernel - mostly stable API. |
49 | |
58 | AnyEvent::MP::Global - stable API. |
50 | stay tuned. |
|
|
51 | |
59 | |
52 | =head1 DESCRIPTION |
60 | =head1 DESCRIPTION |
53 | |
61 | |
54 | This module (-family) implements a simple message passing framework. |
62 | This module (-family) implements a simple message passing framework. |
55 | |
63 | |
56 | Despite its simplicity, you can securely message other processes running |
64 | Despite its simplicity, you can securely message other processes running |
57 | on the same or other hosts. |
65 | on the same or other hosts, and you can supervise entities remotely. |
58 | |
66 | |
59 | For an introduction to this module family, see the L<AnyEvent::MP::Intro> |
67 | For an introduction to this module family, see the L<AnyEvent::MP::Intro> |
60 | manual page. |
68 | manual page and the examples under F<eg/>. |
61 | |
|
|
62 | At the moment, this module family is severly broken and underdocumented, |
|
|
63 | so do not use. This was uploaded mainly to reserve the CPAN namespace - |
|
|
64 | stay tuned! |
|
|
65 | |
69 | |
66 | =head1 CONCEPTS |
70 | =head1 CONCEPTS |
67 | |
71 | |
68 | =over 4 |
72 | =over 4 |
69 | |
73 | |
70 | =item port |
74 | =item port |
71 | |
75 | |
72 | A port is something you can send messages to (with the C<snd> function). |
76 | Not to be confused with a TCP port, a "port" is something you can send |
|
|
77 | messages to (with the C<snd> function). |
73 | |
78 | |
74 | Some ports allow you to register C<rcv> handlers that can match specific |
79 | Ports allow you to register C<rcv> handlers that can match all or just |
75 | messages. All C<rcv> handlers will receive messages they match, messages |
80 | some messages. Messages send to ports will not be queued, regardless of |
76 | will not be queued. |
81 | anything was listening for them or not. |
77 | |
82 | |
78 | =item port id - C<noderef#portname> |
83 | =item port ID - C<nodeid#portname> |
79 | |
84 | |
80 | A port id is normaly the concatenation of a noderef, a hash-mark (C<#>) as |
85 | A port ID is the concatenation of a node ID, a hash-mark (C<#>) as |
81 | separator, and a port name (a printable string of unspecified format). An |
86 | separator, and a port name (a printable string of unspecified format). |
82 | exception is the the node port, whose ID is identical to its node |
|
|
83 | reference. |
|
|
84 | |
87 | |
85 | =item node |
88 | =item node |
86 | |
89 | |
87 | A node is a single process containing at least one port - the node |
90 | A node is a single process containing at least one port - the node port, |
88 | port. You can send messages to node ports to find existing ports or to |
91 | which enables nodes to manage each other remotely, and to create new |
89 | create new ports, among other things. |
92 | ports. |
90 | |
93 | |
91 | Nodes are either private (single-process only), slaves (connected to a |
94 | Nodes are either public (have one or more listening ports) or private |
92 | master node only) or public nodes (connectable from unrelated nodes). |
95 | (no listening ports). Private nodes cannot talk to other private nodes |
|
|
96 | currently. |
93 | |
97 | |
94 | =item noderef - C<host:port,host:port...>, C<id@noderef>, C<id> |
98 | =item node ID - C<[A-Z_][a-zA-Z0-9_\-.:]*> |
95 | |
99 | |
96 | A node reference is a string that either simply identifies the node (for |
100 | A node ID is a string that uniquely identifies the node within a |
97 | private and slave nodes), or contains a recipe on how to reach a given |
101 | network. Depending on the configuration used, node IDs can look like a |
98 | node (for public nodes). |
102 | hostname, a hostname and a port, or a random string. AnyEvent::MP itself |
|
|
103 | doesn't interpret node IDs in any way. |
99 | |
104 | |
100 | This recipe is simply a comma-separated list of C<address:port> pairs (for |
105 | =item binds - C<ip:port> |
101 | TCP/IP, other protocols might look different). |
|
|
102 | |
106 | |
103 | Node references come in two flavours: resolved (containing only numerical |
107 | Nodes can only talk to each other by creating some kind of connection to |
104 | addresses) or unresolved (where hostnames are used instead of addresses). |
108 | each other. To do this, nodes should listen on one or more local transport |
|
|
109 | endpoints - binds. Currently, only standard C<ip:port> specifications can |
|
|
110 | be used, which specify TCP ports to listen on. |
105 | |
111 | |
106 | Before using an unresolved node reference in a message you first have to |
112 | =item seed nodes |
107 | resolve it. |
113 | |
|
|
114 | When a node starts, it knows nothing about the network. To teach the node |
|
|
115 | about the network it first has to contact some other node within the |
|
|
116 | network. This node is called a seed. |
|
|
117 | |
|
|
118 | Apart from the fact that other nodes know them as seed nodes and they have |
|
|
119 | to have fixed listening addresses, seed nodes are perfectly normal nodes - |
|
|
120 | any node can function as a seed node for others. |
|
|
121 | |
|
|
122 | In addition to discovering the network, seed nodes are also used to |
|
|
123 | maintain the network and to connect nodes that otherwise would have |
|
|
124 | trouble connecting. They form the backbone of an AnyEvent::MP network. |
|
|
125 | |
|
|
126 | Seed nodes are expected to be long-running, and at least one seed node |
|
|
127 | should always be available. They should also be relatively responsive - a |
|
|
128 | seed node that blocks for long periods will slow down everybody else. |
|
|
129 | |
|
|
130 | =item seeds - C<host:port> |
|
|
131 | |
|
|
132 | Seeds are transport endpoint(s) (usually a hostname/IP address and a |
|
|
133 | TCP port) of nodes that should be used as seed nodes. |
|
|
134 | |
|
|
135 | The nodes listening on those endpoints are expected to be long-running, |
|
|
136 | and at least one of those should always be available. When nodes run out |
|
|
137 | of connections (e.g. due to a network error), they try to re-establish |
|
|
138 | connections to some seednodes again to join the network. |
108 | |
139 | |
109 | =back |
140 | =back |
110 | |
141 | |
111 | =head1 VARIABLES/FUNCTIONS |
142 | =head1 VARIABLES/FUNCTIONS |
112 | |
143 | |
… | |
… | |
124 | |
155 | |
125 | use AE (); |
156 | use AE (); |
126 | |
157 | |
127 | use base "Exporter"; |
158 | use base "Exporter"; |
128 | |
159 | |
129 | our $VERSION = $AnyEvent::MP::Kernel::VERSION; |
160 | our $VERSION = 1.28; |
130 | |
161 | |
131 | our @EXPORT = qw( |
162 | our @EXPORT = qw( |
132 | NODE $NODE *SELF node_of _any_ |
163 | NODE $NODE *SELF node_of after |
133 | resolve_node initialise_node |
164 | configure |
134 | snd rcv mon kil reg psub spawn |
165 | snd rcv mon mon_guard kil psub peval spawn cal |
135 | port |
166 | port |
136 | ); |
167 | ); |
137 | |
168 | |
138 | our $SELF; |
169 | our $SELF; |
139 | |
170 | |
… | |
… | |
143 | kil $SELF, die => $msg; |
174 | kil $SELF, die => $msg; |
144 | } |
175 | } |
145 | |
176 | |
146 | =item $thisnode = NODE / $NODE |
177 | =item $thisnode = NODE / $NODE |
147 | |
178 | |
148 | The C<NODE> function returns, and the C<$NODE> variable contains the |
179 | The C<NODE> function returns, and the C<$NODE> variable contains, the node |
149 | noderef of the local node. The value is initialised by a call to |
180 | ID of the node running in the current process. This value is initialised by |
150 | C<initialise_node>. |
181 | a call to C<configure>. |
151 | |
182 | |
152 | =item $noderef = node_of $port |
183 | =item $nodeid = node_of $port |
153 | |
184 | |
154 | Extracts and returns the noderef from a port ID or a noderef. |
185 | Extracts and returns the node ID from a port ID or a node ID. |
155 | |
186 | |
156 | =item initialise_node $noderef, $seednode, $seednode... |
187 | =item configure $profile, key => value... |
157 | |
188 | |
158 | =item initialise_node "slave/", $master, $master... |
189 | =item configure key => value... |
159 | |
190 | |
160 | Before a node can talk to other nodes on the network it has to initialise |
191 | Before a node can talk to other nodes on the network (i.e. enter |
161 | itself - the minimum a node needs to know is it's own name, and optionally |
192 | "distributed mode") it has to configure itself - the minimum a node needs |
162 | it should know the noderefs of some other nodes in the network. |
193 | to know is its own name, and optionally it should know the addresses of |
|
|
194 | some other nodes in the network to discover other nodes. |
163 | |
195 | |
|
|
196 | The key/value pairs are basically the same ones as documented for the |
|
|
197 | F<aemp> command line utility (sans the set/del prefix). |
|
|
198 | |
164 | This function initialises a node - it must be called exactly once (or |
199 | This function configures a node - it must be called exactly once (or |
165 | never) before calling other AnyEvent::MP functions. |
200 | never) before calling other AnyEvent::MP functions. |
166 | |
201 | |
167 | All arguments (optionally except for the first) are noderefs, which can be |
|
|
168 | either resolved or unresolved. |
|
|
169 | |
|
|
170 | The first argument will be looked up in the configuration database first |
|
|
171 | (if it is C<undef> then the current nodename will be used instead) to find |
|
|
172 | the relevant configuration profile (see L<aemp>). If none is found then |
|
|
173 | the default configuration is used. The configuration supplies additional |
|
|
174 | seed/master nodes and can override the actual noderef. |
|
|
175 | |
|
|
176 | There are two types of networked nodes, public nodes and slave nodes: |
|
|
177 | |
|
|
178 | =over 4 |
202 | =over 4 |
179 | |
203 | |
180 | =item public nodes |
204 | =item step 1, gathering configuration from profiles |
181 | |
205 | |
182 | For public nodes, C<$noderef> (supplied either directly to |
206 | The function first looks up a profile in the aemp configuration (see the |
183 | C<initialise_node> or indirectly via a profile or the nodename) must be a |
207 | L<aemp> commandline utility). The profile name can be specified via the |
184 | noderef (possibly unresolved, in which case it will be resolved). |
208 | named C<profile> parameter or can simply be the first parameter). If it is |
|
|
209 | missing, then the nodename (F<uname -n>) will be used as profile name. |
185 | |
210 | |
186 | After resolving, the node will bind itself on all endpoints and try to |
211 | The profile data is then gathered as follows: |
187 | connect to all additional C<$seednodes> that are specified. Seednodes are |
|
|
188 | optional and can be used to quickly bootstrap the node into an existing |
|
|
189 | network. |
|
|
190 | |
212 | |
191 | =item slave nodes |
213 | First, all remaining key => value pairs (all of which are conveniently |
|
|
214 | undocumented at the moment) will be interpreted as configuration |
|
|
215 | data. Then they will be overwritten by any values specified in the global |
|
|
216 | default configuration (see the F<aemp> utility), then the chain of |
|
|
217 | profiles chosen by the profile name (and any C<parent> attributes). |
192 | |
218 | |
193 | When the C<$noderef> (either as given or overriden by the config file) |
219 | That means that the values specified in the profile have highest priority |
194 | is the special string C<slave/>, then the node will become a slave |
220 | and the values specified directly via C<configure> have lowest priority, |
195 | node. Slave nodes cannot be contacted from outside and will route most of |
221 | and can only be used to specify defaults. |
196 | their traffic to the master node that they attach to. |
|
|
197 | |
222 | |
198 | At least one additional noderef is required (either by specifying it |
223 | If the profile specifies a node ID, then this will become the node ID of |
199 | directly or because it is part of the configuration profile): The node |
224 | this process. If not, then the profile name will be used as node ID. The |
200 | will try to connect to all of them and will become a slave attached to the |
225 | special node ID of C<anon/> will be replaced by a random node ID. |
201 | first node it can successfully connect to. |
226 | |
|
|
227 | =item step 2, bind listener sockets |
|
|
228 | |
|
|
229 | The next step is to look up the binds in the profile, followed by binding |
|
|
230 | aemp protocol listeners on all binds specified (it is possible and valid |
|
|
231 | to have no binds, meaning that the node cannot be contacted form the |
|
|
232 | outside. This means the node cannot talk to other nodes that also have no |
|
|
233 | binds, but it can still talk to all "normal" nodes). |
|
|
234 | |
|
|
235 | If the profile does not specify a binds list, then a default of C<*> is |
|
|
236 | used, meaning the node will bind on a dynamically-assigned port on every |
|
|
237 | local IP address it finds. |
|
|
238 | |
|
|
239 | =item step 3, connect to seed nodes |
|
|
240 | |
|
|
241 | As the last step, the seeds list from the profile is passed to the |
|
|
242 | L<AnyEvent::MP::Global> module, which will then use it to keep |
|
|
243 | connectivity with at least one node at any point in time. |
202 | |
244 | |
203 | =back |
245 | =back |
204 | |
246 | |
205 | This function will block until all nodes have been resolved and, for slave |
247 | Example: become a distributed node using the local node name as profile. |
206 | nodes, until it has successfully established a connection to a master |
248 | This should be the most common form of invocation for "daemon"-type nodes. |
207 | server. |
|
|
208 | |
249 | |
209 | Example: become a public node listening on the guessed noderef, or the one |
250 | configure |
210 | specified via C<aemp> for the current node. This should be the most common |
|
|
211 | form of invocation for "daemon"-type nodes. |
|
|
212 | |
251 | |
213 | initialise_node; |
252 | Example: become an anonymous node. This form is often used for commandline |
|
|
253 | clients. |
214 | |
254 | |
215 | Example: become a slave node to any of the the seednodes specified via |
255 | configure nodeid => "anon/"; |
216 | C<aemp>. This form is often used for commandline clients. |
|
|
217 | |
256 | |
218 | initialise_node "slave/"; |
257 | Example: configure a node using a profile called seed, which si suitable |
|
|
258 | for a seed node as it binds on all local addresses on a fixed port (4040, |
|
|
259 | customary for aemp). |
219 | |
260 | |
220 | Example: become a slave node to any of the specified master servers. This |
261 | # use the aemp commandline utility |
221 | form is also often used for commandline clients. |
262 | # aemp profile seed nodeid anon/ binds '*:4040' |
222 | |
263 | |
223 | initialise_node "slave/", "master1", "192.168.13.17", "mp.example.net"; |
264 | # then use it |
|
|
265 | configure profile => "seed"; |
224 | |
266 | |
225 | Example: become a public node, and try to contact some well-known master |
267 | # or simply use aemp from the shell again: |
226 | servers to become part of the network. |
268 | # aemp run profile seed |
227 | |
269 | |
228 | initialise_node undef, "master1", "master2"; |
270 | # or provide a nicer-to-remember nodeid |
229 | |
271 | # aemp run profile seed nodeid "$(hostname)" |
230 | Example: become a public node listening on port C<4041>. |
|
|
231 | |
|
|
232 | initialise_node 4041; |
|
|
233 | |
|
|
234 | Example: become a public node, only visible on localhost port 4044. |
|
|
235 | |
|
|
236 | initialise_node "localhost:4044"; |
|
|
237 | |
|
|
238 | =item $cv = resolve_node $noderef |
|
|
239 | |
|
|
240 | Takes an unresolved node reference that may contain hostnames and |
|
|
241 | abbreviated IDs, resolves all of them and returns a resolved node |
|
|
242 | reference. |
|
|
243 | |
|
|
244 | In addition to C<address:port> pairs allowed in resolved noderefs, the |
|
|
245 | following forms are supported: |
|
|
246 | |
|
|
247 | =over 4 |
|
|
248 | |
|
|
249 | =item the empty string |
|
|
250 | |
|
|
251 | An empty-string component gets resolved as if the default port (4040) was |
|
|
252 | specified. |
|
|
253 | |
|
|
254 | =item naked port numbers (e.g. C<1234>) |
|
|
255 | |
|
|
256 | These are resolved by prepending the local nodename and a colon, to be |
|
|
257 | further resolved. |
|
|
258 | |
|
|
259 | =item hostnames (e.g. C<localhost:1234>, C<localhost>) |
|
|
260 | |
|
|
261 | These are resolved by using AnyEvent::DNS to resolve them, optionally |
|
|
262 | looking up SRV records for the C<aemp=4040> port, if no port was |
|
|
263 | specified. |
|
|
264 | |
|
|
265 | =back |
|
|
266 | |
272 | |
267 | =item $SELF |
273 | =item $SELF |
268 | |
274 | |
269 | Contains the current port id while executing C<rcv> callbacks or C<psub> |
275 | Contains the current port id while executing C<rcv> callbacks or C<psub> |
270 | blocks. |
276 | blocks. |
271 | |
277 | |
272 | =item SELF, %SELF, @SELF... |
278 | =item *SELF, SELF, %SELF, @SELF... |
273 | |
279 | |
274 | Due to some quirks in how perl exports variables, it is impossible to |
280 | Due to some quirks in how perl exports variables, it is impossible to |
275 | just export C<$SELF>, all the symbols called C<SELF> are exported by this |
281 | just export C<$SELF>, all the symbols named C<SELF> are exported by this |
276 | module, but only C<$SELF> is currently used. |
282 | module, but only C<$SELF> is currently used. |
277 | |
283 | |
278 | =item snd $port, type => @data |
284 | =item snd $port, type => @data |
279 | |
285 | |
280 | =item snd $port, @msg |
286 | =item snd $port, @msg |
281 | |
287 | |
282 | Send the given message to the given port ID, which can identify either |
288 | Send the given message to the given port, which can identify either a |
283 | a local or a remote port, and must be a port ID. |
289 | local or a remote port, and must be a port ID. |
284 | |
290 | |
285 | While the message can be about anything, it is highly recommended to use a |
291 | While the message can be almost anything, it is highly recommended to |
286 | string as first element (a port ID, or some word that indicates a request |
292 | use a string as first element (a port ID, or some word that indicates a |
287 | type etc.). |
293 | request type etc.) and to consist if only simple perl values (scalars, |
|
|
294 | arrays, hashes) - if you think you need to pass an object, think again. |
288 | |
295 | |
289 | The message data effectively becomes read-only after a call to this |
296 | The message data logically becomes read-only after a call to this |
290 | function: modifying any argument is not allowed and can cause many |
297 | function: modifying any argument (or values referenced by them) is |
291 | problems. |
298 | forbidden, as there can be considerable time between the call to C<snd> |
|
|
299 | and the time the message is actually being serialised - in fact, it might |
|
|
300 | never be copied as within the same process it is simply handed to the |
|
|
301 | receiving port. |
292 | |
302 | |
293 | The type of data you can transfer depends on the transport protocol: when |
303 | The type of data you can transfer depends on the transport protocol: when |
294 | JSON is used, then only strings, numbers and arrays and hashes consisting |
304 | JSON is used, then only strings, numbers and arrays and hashes consisting |
295 | of those are allowed (no objects). When Storable is used, then anything |
305 | of those are allowed (no objects). When Storable is used, then anything |
296 | that Storable can serialise and deserialise is allowed, and for the local |
306 | that Storable can serialise and deserialise is allowed, and for the local |
297 | node, anything can be passed. |
307 | node, anything can be passed. Best rely only on the common denominator of |
|
|
308 | these. |
298 | |
309 | |
299 | =item $local_port = port |
310 | =item $local_port = port |
300 | |
311 | |
301 | Create a new local port object and returns its port ID. Initially it has |
312 | Create a new local port object and returns its port ID. Initially it has |
302 | no callbacks set and will throw an error when it receives messages. |
313 | no callbacks set and will throw an error when it receives messages. |
… | |
… | |
349 | The default callback received all messages not matched by a more specific |
360 | The default callback received all messages not matched by a more specific |
350 | C<tag> match. |
361 | C<tag> match. |
351 | |
362 | |
352 | =item rcv $local_port, tag => $callback->(@msg_without_tag), ... |
363 | =item rcv $local_port, tag => $callback->(@msg_without_tag), ... |
353 | |
364 | |
354 | Register callbacks to be called on messages starting with the given tag on |
365 | Register (or replace) callbacks to be called on messages starting with the |
355 | the given port (and return the port), or unregister it (when C<$callback> |
366 | given tag on the given port (and return the port), or unregister it (when |
356 | is C<$undef>). |
367 | C<$callback> is C<$undef> or missing). There can only be one callback |
|
|
368 | registered for each tag. |
357 | |
369 | |
358 | The original message will be passed to the callback, after the first |
370 | The original message will be passed to the callback, after the first |
359 | element (the tag) has been removed. The callback will use the same |
371 | element (the tag) has been removed. The callback will use the same |
360 | environment as the default callback (see above). |
372 | environment as the default callback (see above). |
361 | |
373 | |
… | |
… | |
373 | rcv port, |
385 | rcv port, |
374 | msg1 => sub { ... }, |
386 | msg1 => sub { ... }, |
375 | ... |
387 | ... |
376 | ; |
388 | ; |
377 | |
389 | |
|
|
390 | Example: temporarily register a rcv callback for a tag matching some port |
|
|
391 | (e.g. for an rpc reply) and unregister it after a message was received. |
|
|
392 | |
|
|
393 | rcv $port, $otherport => sub { |
|
|
394 | my @reply = @_; |
|
|
395 | |
|
|
396 | rcv $SELF, $otherport; |
|
|
397 | }; |
|
|
398 | |
378 | =cut |
399 | =cut |
379 | |
400 | |
380 | sub rcv($@) { |
401 | sub rcv($@) { |
381 | my $port = shift; |
402 | my $port = shift; |
382 | my ($noderef, $portid) = split /#/, $port, 2; |
403 | my ($nodeid, $portid) = split /#/, $port, 2; |
383 | |
404 | |
384 | ($NODE{$noderef} || add_node $noderef) == $NODE{""} |
405 | $NODE{$nodeid} == $NODE{""} |
385 | or Carp::croak "$port: rcv can only be called on local ports, caught"; |
406 | or Carp::croak "$port: rcv can only be called on local ports, caught"; |
386 | |
407 | |
387 | while (@_) { |
408 | while (@_) { |
388 | if (ref $_[0]) { |
409 | if (ref $_[0]) { |
389 | if (my $self = $PORT_DATA{$portid}) { |
410 | if (my $self = $PORT_DATA{$portid}) { |
390 | "AnyEvent::MP::Port" eq ref $self |
411 | "AnyEvent::MP::Port" eq ref $self |
391 | or Carp::croak "$port: rcv can only be called on message matching ports, caught"; |
412 | or Carp::croak "$port: rcv can only be called on message matching ports, caught"; |
392 | |
413 | |
393 | $self->[2] = shift; |
414 | $self->[0] = shift; |
394 | } else { |
415 | } else { |
395 | my $cb = shift; |
416 | my $cb = shift; |
396 | $PORT{$portid} = sub { |
417 | $PORT{$portid} = sub { |
397 | local $SELF = $port; |
418 | local $SELF = $port; |
398 | eval { &$cb }; _self_die if $@; |
419 | eval { &$cb }; _self_die if $@; |
399 | }; |
420 | }; |
400 | } |
421 | } |
401 | } elsif (defined $_[0]) { |
422 | } elsif (defined $_[0]) { |
402 | my $self = $PORT_DATA{$portid} ||= do { |
423 | my $self = $PORT_DATA{$portid} ||= do { |
403 | my $self = bless [$PORT{$port} || sub { }, { }, $port], "AnyEvent::MP::Port"; |
424 | my $self = bless [$PORT{$portid} || sub { }, { }, $port], "AnyEvent::MP::Port"; |
404 | |
425 | |
405 | $PORT{$portid} = sub { |
426 | $PORT{$portid} = sub { |
406 | local $SELF = $port; |
427 | local $SELF = $port; |
407 | |
428 | |
408 | if (my $cb = $self->[1]{$_[0]}) { |
429 | if (my $cb = $self->[1]{$_[0]}) { |
… | |
… | |
430 | } |
451 | } |
431 | |
452 | |
432 | $port |
453 | $port |
433 | } |
454 | } |
434 | |
455 | |
|
|
456 | =item peval $port, $coderef[, @args] |
|
|
457 | |
|
|
458 | Evaluates the given C<$codref> within the contetx of C<$port>, that is, |
|
|
459 | when the code throews an exception the C<$port> will be killed. |
|
|
460 | |
|
|
461 | Any remaining args will be passed to the callback. Any return values will |
|
|
462 | be returned to the caller. |
|
|
463 | |
|
|
464 | This is useful when you temporarily want to execute code in the context of |
|
|
465 | a port. |
|
|
466 | |
|
|
467 | Example: create a port and run some initialisation code in it's context. |
|
|
468 | |
|
|
469 | my $port = port { ... }; |
|
|
470 | |
|
|
471 | peval $port, sub { |
|
|
472 | init |
|
|
473 | or die "unable to init"; |
|
|
474 | }; |
|
|
475 | |
|
|
476 | =cut |
|
|
477 | |
|
|
478 | sub peval($$) { |
|
|
479 | local $SELF = shift; |
|
|
480 | my $cb = shift; |
|
|
481 | |
|
|
482 | if (wantarray) { |
|
|
483 | my @res = eval { &$cb }; |
|
|
484 | _self_die if $@; |
|
|
485 | @res |
|
|
486 | } else { |
|
|
487 | my $res = eval { &$cb }; |
|
|
488 | _self_die if $@; |
|
|
489 | $res |
|
|
490 | } |
|
|
491 | } |
|
|
492 | |
435 | =item $closure = psub { BLOCK } |
493 | =item $closure = psub { BLOCK } |
436 | |
494 | |
437 | Remembers C<$SELF> and creates a closure out of the BLOCK. When the |
495 | Remembers C<$SELF> and creates a closure out of the BLOCK. When the |
438 | closure is executed, sets up the environment in the same way as in C<rcv> |
496 | closure is executed, sets up the environment in the same way as in C<rcv> |
439 | callbacks, i.e. runtime errors will cause the port to get C<kil>ed. |
497 | callbacks, i.e. runtime errors will cause the port to get C<kil>ed. |
|
|
498 | |
|
|
499 | The effect is basically as if it returned C<< sub { peval $SELF, sub { |
|
|
500 | BLOCK } } >>. |
440 | |
501 | |
441 | This is useful when you register callbacks from C<rcv> callbacks: |
502 | This is useful when you register callbacks from C<rcv> callbacks: |
442 | |
503 | |
443 | rcv delayed_reply => sub { |
504 | rcv delayed_reply => sub { |
444 | my ($delay, @reply) = @_; |
505 | my ($delay, @reply) = @_; |
… | |
… | |
468 | $res |
529 | $res |
469 | } |
530 | } |
470 | } |
531 | } |
471 | } |
532 | } |
472 | |
533 | |
473 | =item $guard = mon $port, $cb->(@reason) |
534 | =item $guard = mon $port, $cb->(@reason) # call $cb when $port dies |
474 | |
535 | |
475 | =item $guard = mon $port, $rcvport |
536 | =item $guard = mon $port, $rcvport # kill $rcvport when $port dies |
476 | |
537 | |
477 | =item $guard = mon $port |
538 | =item $guard = mon $port # kill $SELF when $port dies |
478 | |
539 | |
479 | =item $guard = mon $port, $rcvport, @msg |
540 | =item $guard = mon $port, $rcvport, @msg # send a message when $port dies |
480 | |
541 | |
481 | Monitor the given port and do something when the port is killed or |
542 | Monitor the given port and do something when the port is killed or |
482 | messages to it were lost, and optionally return a guard that can be used |
543 | messages to it were lost, and optionally return a guard that can be used |
483 | to stop monitoring again. |
544 | to stop monitoring again. |
484 | |
|
|
485 | C<mon> effectively guarantees that, in the absence of hardware failures, |
|
|
486 | that after starting the monitor, either all messages sent to the port |
|
|
487 | will arrive, or the monitoring action will be invoked after possible |
|
|
488 | message loss has been detected. No messages will be lost "in between" |
|
|
489 | (after the first lost message no further messages will be received by the |
|
|
490 | port). After the monitoring action was invoked, further messages might get |
|
|
491 | delivered again. |
|
|
492 | |
545 | |
493 | In the first form (callback), the callback is simply called with any |
546 | In the first form (callback), the callback is simply called with any |
494 | number of C<@reason> elements (no @reason means that the port was deleted |
547 | number of C<@reason> elements (no @reason means that the port was deleted |
495 | "normally"). Note also that I<< the callback B<must> never die >>, so use |
548 | "normally"). Note also that I<< the callback B<must> never die >>, so use |
496 | C<eval> if unsure. |
549 | C<eval> if unsure. |
497 | |
550 | |
498 | In the second form (another port given), the other port (C<$rcvport>) |
551 | In the second form (another port given), the other port (C<$rcvport>) |
499 | will be C<kil>'ed with C<@reason>, iff a @reason was specified, i.e. on |
552 | will be C<kil>'ed with C<@reason>, if a @reason was specified, i.e. on |
500 | "normal" kils nothing happens, while under all other conditions, the other |
553 | "normal" kils nothing happens, while under all other conditions, the other |
501 | port is killed with the same reason. |
554 | port is killed with the same reason. |
502 | |
555 | |
503 | The third form (kill self) is the same as the second form, except that |
556 | The third form (kill self) is the same as the second form, except that |
504 | C<$rvport> defaults to C<$SELF>. |
557 | C<$rvport> defaults to C<$SELF>. |
505 | |
558 | |
506 | In the last form (message), a message of the form C<@msg, @reason> will be |
559 | In the last form (message), a message of the form C<@msg, @reason> will be |
507 | C<snd>. |
560 | C<snd>. |
|
|
561 | |
|
|
562 | Monitoring-actions are one-shot: once messages are lost (and a monitoring |
|
|
563 | alert was raised), they are removed and will not trigger again. |
508 | |
564 | |
509 | As a rule of thumb, monitoring requests should always monitor a port from |
565 | As a rule of thumb, monitoring requests should always monitor a port from |
510 | a local port (or callback). The reason is that kill messages might get |
566 | a local port (or callback). The reason is that kill messages might get |
511 | lost, just like any other message. Another less obvious reason is that |
567 | lost, just like any other message. Another less obvious reason is that |
512 | even monitoring requests can get lost (for exmaple, when the connection |
568 | even monitoring requests can get lost (for example, when the connection |
513 | to the other node goes down permanently). When monitoring a port locally |
569 | to the other node goes down permanently). When monitoring a port locally |
514 | these problems do not exist. |
570 | these problems do not exist. |
515 | |
571 | |
|
|
572 | C<mon> effectively guarantees that, in the absence of hardware failures, |
|
|
573 | after starting the monitor, either all messages sent to the port will |
|
|
574 | arrive, or the monitoring action will be invoked after possible message |
|
|
575 | loss has been detected. No messages will be lost "in between" (after |
|
|
576 | the first lost message no further messages will be received by the |
|
|
577 | port). After the monitoring action was invoked, further messages might get |
|
|
578 | delivered again. |
|
|
579 | |
|
|
580 | Inter-host-connection timeouts and monitoring depend on the transport |
|
|
581 | used. The only transport currently implemented is TCP, and AnyEvent::MP |
|
|
582 | relies on TCP to detect node-downs (this can take 10-15 minutes on a |
|
|
583 | non-idle connection, and usually around two hours for idle connections). |
|
|
584 | |
|
|
585 | This means that monitoring is good for program errors and cleaning up |
|
|
586 | stuff eventually, but they are no replacement for a timeout when you need |
|
|
587 | to ensure some maximum latency. |
|
|
588 | |
516 | Example: call a given callback when C<$port> is killed. |
589 | Example: call a given callback when C<$port> is killed. |
517 | |
590 | |
518 | mon $port, sub { warn "port died because of <@_>\n" }; |
591 | mon $port, sub { warn "port died because of <@_>\n" }; |
519 | |
592 | |
520 | Example: kill ourselves when C<$port> is killed abnormally. |
593 | Example: kill ourselves when C<$port> is killed abnormally. |
… | |
… | |
526 | mon $port, $self => "restart"; |
599 | mon $port, $self => "restart"; |
527 | |
600 | |
528 | =cut |
601 | =cut |
529 | |
602 | |
530 | sub mon { |
603 | sub mon { |
531 | my ($noderef, $port) = split /#/, shift, 2; |
604 | my ($nodeid, $port) = split /#/, shift, 2; |
532 | |
605 | |
533 | my $node = $NODE{$noderef} || add_node $noderef; |
606 | my $node = $NODE{$nodeid} || add_node $nodeid; |
534 | |
607 | |
535 | my $cb = @_ ? shift : $SELF || Carp::croak 'mon: called with one argument only, but $SELF not set,'; |
608 | my $cb = @_ ? shift : $SELF || Carp::croak 'mon: called with one argument only, but $SELF not set,'; |
536 | |
609 | |
537 | unless (ref $cb) { |
610 | unless (ref $cb) { |
538 | if (@_) { |
611 | if (@_) { |
… | |
… | |
547 | } |
620 | } |
548 | |
621 | |
549 | $node->monitor ($port, $cb); |
622 | $node->monitor ($port, $cb); |
550 | |
623 | |
551 | defined wantarray |
624 | defined wantarray |
552 | and AnyEvent::Util::guard { $node->unmonitor ($port, $cb) } |
625 | and ($cb += 0, AnyEvent::Util::guard { $node->unmonitor ($port, $cb) }) |
553 | } |
626 | } |
554 | |
627 | |
555 | =item $guard = mon_guard $port, $ref, $ref... |
628 | =item $guard = mon_guard $port, $ref, $ref... |
556 | |
629 | |
557 | Monitors the given C<$port> and keeps the passed references. When the port |
630 | Monitors the given C<$port> and keeps the passed references. When the port |
558 | is killed, the references will be freed. |
631 | is killed, the references will be freed. |
559 | |
632 | |
560 | Optionally returns a guard that will stop the monitoring. |
633 | Optionally returns a guard that will stop the monitoring. |
561 | |
634 | |
562 | This function is useful when you create e.g. timers or other watchers and |
635 | This function is useful when you create e.g. timers or other watchers and |
563 | want to free them when the port gets killed: |
636 | want to free them when the port gets killed (note the use of C<psub>): |
564 | |
637 | |
565 | $port->rcv (start => sub { |
638 | $port->rcv (start => sub { |
566 | my $timer; $timer = mon_guard $port, AE::timer 1, 1, sub { |
639 | my $timer; $timer = mon_guard $port, AE::timer 1, 1, psub { |
567 | undef $timer if 0.9 < rand; |
640 | undef $timer if 0.9 < rand; |
568 | }); |
641 | }); |
569 | }); |
642 | }); |
570 | |
643 | |
571 | =cut |
644 | =cut |
… | |
… | |
580 | |
653 | |
581 | =item kil $port[, @reason] |
654 | =item kil $port[, @reason] |
582 | |
655 | |
583 | Kill the specified port with the given C<@reason>. |
656 | Kill the specified port with the given C<@reason>. |
584 | |
657 | |
585 | If no C<@reason> is specified, then the port is killed "normally" (linked |
658 | If no C<@reason> is specified, then the port is killed "normally" - |
586 | ports will not be kileld, or even notified). |
659 | monitor callback will be invoked, but the kil will not cause linked ports |
|
|
660 | (C<mon $mport, $lport> form) to get killed. |
587 | |
661 | |
588 | Otherwise, linked ports get killed with the same reason (second form of |
662 | If a C<@reason> is specified, then linked ports (C<mon $mport, $lport> |
589 | C<mon>, see below). |
663 | form) get killed with the same reason. |
590 | |
664 | |
591 | Runtime errors while evaluating C<rcv> callbacks or inside C<psub> blocks |
665 | Runtime errors while evaluating C<rcv> callbacks or inside C<psub> blocks |
592 | will be reported as reason C<< die => $@ >>. |
666 | will be reported as reason C<< die => $@ >>. |
593 | |
667 | |
594 | Transport/communication errors are reported as C<< transport_error => |
668 | Transport/communication errors are reported as C<< transport_error => |
… | |
… | |
599 | =item $port = spawn $node, $initfunc[, @initdata] |
673 | =item $port = spawn $node, $initfunc[, @initdata] |
600 | |
674 | |
601 | Creates a port on the node C<$node> (which can also be a port ID, in which |
675 | Creates a port on the node C<$node> (which can also be a port ID, in which |
602 | case it's the node where that port resides). |
676 | case it's the node where that port resides). |
603 | |
677 | |
604 | The port ID of the newly created port is return immediately, and it is |
678 | The port ID of the newly created port is returned immediately, and it is |
605 | permissible to immediately start sending messages or monitor the port. |
679 | possible to immediately start sending messages or to monitor the port. |
606 | |
680 | |
607 | After the port has been created, the init function is |
681 | After the port has been created, the init function is called on the remote |
608 | called. This function must be a fully-qualified function name |
682 | node, in the same context as a C<rcv> callback. This function must be a |
609 | (e.g. C<MyApp::Chat::Server::init>). To specify a function in the main |
683 | fully-qualified function name (e.g. C<MyApp::Chat::Server::init>). To |
610 | program, use C<::name>. |
684 | specify a function in the main program, use C<::name>. |
611 | |
685 | |
612 | If the function doesn't exist, then the node tries to C<require> |
686 | If the function doesn't exist, then the node tries to C<require> |
613 | the package, then the package above the package and so on (e.g. |
687 | the package, then the package above the package and so on (e.g. |
614 | C<MyApp::Chat::Server>, C<MyApp::Chat>, C<MyApp>) until the function |
688 | C<MyApp::Chat::Server>, C<MyApp::Chat>, C<MyApp>) until the function |
615 | exists or it runs out of package names. |
689 | exists or it runs out of package names. |
616 | |
690 | |
617 | The init function is then called with the newly-created port as context |
691 | The init function is then called with the newly-created port as context |
618 | object (C<$SELF>) and the C<@initdata> values as arguments. |
692 | object (C<$SELF>) and the C<@initdata> values as arguments. It I<must> |
|
|
693 | call one of the C<rcv> functions to set callbacks on C<$SELF>, otherwise |
|
|
694 | the port might not get created. |
619 | |
695 | |
620 | A common idiom is to pass your own port, monitor the spawned port, and |
696 | A common idiom is to pass a local port, immediately monitor the spawned |
621 | in the init function, monitor the original port. This two-way monitoring |
697 | port, and in the remote init function, immediately monitor the passed |
622 | ensures that both ports get cleaned up when there is a problem. |
698 | local port. This two-way monitoring ensures that both ports get cleaned up |
|
|
699 | when there is a problem. |
|
|
700 | |
|
|
701 | C<spawn> guarantees that the C<$initfunc> has no visible effects on the |
|
|
702 | caller before C<spawn> returns (by delaying invocation when spawn is |
|
|
703 | called for the local node). |
623 | |
704 | |
624 | Example: spawn a chat server port on C<$othernode>. |
705 | Example: spawn a chat server port on C<$othernode>. |
625 | |
706 | |
626 | # this node, executed from within a port context: |
707 | # this node, executed from within a port context: |
627 | my $server = spawn $othernode, "MyApp::Chat::Server::connect", $SELF; |
708 | my $server = spawn $othernode, "MyApp::Chat::Server::connect", $SELF; |
… | |
… | |
642 | |
723 | |
643 | sub _spawn { |
724 | sub _spawn { |
644 | my $port = shift; |
725 | my $port = shift; |
645 | my $init = shift; |
726 | my $init = shift; |
646 | |
727 | |
|
|
728 | # rcv will create the actual port |
647 | local $SELF = "$NODE#$port"; |
729 | local $SELF = "$NODE#$port"; |
648 | eval { |
730 | eval { |
649 | &{ load_func $init } |
731 | &{ load_func $init } |
650 | }; |
732 | }; |
651 | _self_die if $@; |
733 | _self_die if $@; |
652 | } |
734 | } |
653 | |
735 | |
654 | sub spawn(@) { |
736 | sub spawn(@) { |
655 | my ($noderef, undef) = split /#/, shift, 2; |
737 | my ($nodeid, undef) = split /#/, shift, 2; |
656 | |
738 | |
657 | my $id = "$RUNIQ." . $ID++; |
739 | my $id = "$RUNIQ." . $ID++; |
658 | |
740 | |
659 | $_[0] =~ /::/ |
741 | $_[0] =~ /::/ |
660 | or Carp::croak "spawn init function must be a fully-qualified name, caught"; |
742 | or Carp::croak "spawn init function must be a fully-qualified name, caught"; |
661 | |
743 | |
662 | ($NODE{$noderef} || add_node $noderef) |
744 | snd_to_func $nodeid, "AnyEvent::MP::_spawn" => $id, @_; |
663 | ->send (["", "AnyEvent::MP::_spawn" => $id, @_]); |
|
|
664 | |
745 | |
665 | "$noderef#$id" |
746 | "$nodeid#$id" |
666 | } |
747 | } |
667 | |
748 | |
668 | =back |
749 | =item after $timeout, @msg |
669 | |
750 | |
670 | =head1 NODE MESSAGES |
751 | =item after $timeout, $callback |
671 | |
752 | |
672 | Nodes understand the following messages sent to them. Many of them take |
753 | Either sends the given message, or call the given callback, after the |
673 | arguments called C<@reply>, which will simply be used to compose a reply |
754 | specified number of seconds. |
674 | message - C<$reply[0]> is the port to reply to, C<$reply[1]> the type and |
|
|
675 | the remaining arguments are simply the message data. |
|
|
676 | |
755 | |
677 | While other messages exist, they are not public and subject to change. |
756 | This is simply a utility function that comes in handy at times - the |
|
|
757 | AnyEvent::MP author is not convinced of the wisdom of having it, though, |
|
|
758 | so it may go away in the future. |
678 | |
759 | |
679 | =over 4 |
|
|
680 | |
|
|
681 | =cut |
760 | =cut |
682 | |
761 | |
683 | =item lookup => $name, @reply |
762 | sub after($@) { |
|
|
763 | my ($timeout, @action) = @_; |
684 | |
764 | |
685 | Replies with the port ID of the specified well-known port, or C<undef>. |
765 | my $t; $t = AE::timer $timeout, 0, sub { |
|
|
766 | undef $t; |
|
|
767 | ref $action[0] |
|
|
768 | ? $action[0]() |
|
|
769 | : snd @action; |
|
|
770 | }; |
|
|
771 | } |
686 | |
772 | |
687 | =item devnull => ... |
773 | =item cal $port, @msg, $callback[, $timeout] |
688 | |
774 | |
689 | Generic data sink/CPU heat conversion. |
775 | A simple form of RPC - sends a message to the given C<$port> with the |
|
|
776 | given contents (C<@msg>), but adds a reply port to the message. |
690 | |
777 | |
691 | =item relay => $port, @msg |
778 | The reply port is created temporarily just for the purpose of receiving |
|
|
779 | the reply, and will be C<kil>ed when no longer needed. |
692 | |
780 | |
693 | Simply forwards the message to the given port. |
781 | A reply message sent to the port is passed to the C<$callback> as-is. |
694 | |
782 | |
695 | =item eval => $string[ @reply] |
783 | If an optional time-out (in seconds) is given and it is not C<undef>, |
|
|
784 | then the callback will be called without any arguments after the time-out |
|
|
785 | elapsed and the port is C<kil>ed. |
696 | |
786 | |
697 | Evaluates the given string. If C<@reply> is given, then a message of the |
787 | If no time-out is given (or it is C<undef>), then the local port will |
698 | form C<@reply, $@, @evalres> is sent. |
788 | monitor the remote port instead, so it eventually gets cleaned-up. |
699 | |
789 | |
700 | Example: crash another node. |
790 | Currently this function returns the temporary port, but this "feature" |
|
|
791 | might go in future versions unless you can make a convincing case that |
|
|
792 | this is indeed useful for something. |
701 | |
793 | |
702 | snd $othernode, eval => "exit"; |
794 | =cut |
703 | |
795 | |
704 | =item time => @reply |
796 | sub cal(@) { |
|
|
797 | my $timeout = ref $_[-1] ? undef : pop; |
|
|
798 | my $cb = pop; |
705 | |
799 | |
706 | Replies the the current node time to C<@reply>. |
800 | my $port = port { |
|
|
801 | undef $timeout; |
|
|
802 | kil $SELF; |
|
|
803 | &$cb; |
|
|
804 | }; |
707 | |
805 | |
708 | Example: tell the current node to send the current time to C<$myport> in a |
806 | if (defined $timeout) { |
709 | C<timereply> message. |
807 | $timeout = AE::timer $timeout, 0, sub { |
|
|
808 | undef $timeout; |
|
|
809 | kil $port; |
|
|
810 | $cb->(); |
|
|
811 | }; |
|
|
812 | } else { |
|
|
813 | mon $_[0], sub { |
|
|
814 | kil $port; |
|
|
815 | $cb->(); |
|
|
816 | }; |
|
|
817 | } |
710 | |
818 | |
711 | snd $NODE, time => $myport, timereply => 1, 2; |
819 | push @_, $port; |
712 | # => snd $myport, timereply => 1, 2, <time> |
820 | &snd; |
|
|
821 | |
|
|
822 | $port |
|
|
823 | } |
713 | |
824 | |
714 | =back |
825 | =back |
715 | |
826 | |
716 | =head1 AnyEvent::MP vs. Distributed Erlang |
827 | =head1 AnyEvent::MP vs. Distributed Erlang |
717 | |
828 | |
718 | AnyEvent::MP got lots of its ideas from distributed Erlang (Erlang node |
829 | AnyEvent::MP got lots of its ideas from distributed Erlang (Erlang node |
719 | == aemp node, Erlang process == aemp port), so many of the documents and |
830 | == aemp node, Erlang process == aemp port), so many of the documents and |
720 | programming techniques employed by Erlang apply to AnyEvent::MP. Here is a |
831 | programming techniques employed by Erlang apply to AnyEvent::MP. Here is a |
721 | sample: |
832 | sample: |
722 | |
833 | |
723 | http://www.Erlang.se/doc/programming_rules.shtml |
834 | http://www.erlang.se/doc/programming_rules.shtml |
724 | http://Erlang.org/doc/getting_started/part_frame.html # chapters 3 and 4 |
835 | http://erlang.org/doc/getting_started/part_frame.html # chapters 3 and 4 |
725 | http://Erlang.org/download/Erlang-book-part1.pdf # chapters 5 and 6 |
836 | http://erlang.org/download/erlang-book-part1.pdf # chapters 5 and 6 |
726 | http://Erlang.org/download/armstrong_thesis_2003.pdf # chapters 4 and 5 |
837 | http://erlang.org/download/armstrong_thesis_2003.pdf # chapters 4 and 5 |
727 | |
838 | |
728 | Despite the similarities, there are also some important differences: |
839 | Despite the similarities, there are also some important differences: |
729 | |
840 | |
730 | =over 4 |
841 | =over 4 |
731 | |
842 | |
732 | =item * Node references contain the recipe on how to contact them. |
843 | =item * Node IDs are arbitrary strings in AEMP. |
733 | |
844 | |
734 | Erlang relies on special naming and DNS to work everywhere in the |
845 | Erlang relies on special naming and DNS to work everywhere in the same |
735 | same way. AEMP relies on each node knowing it's own address(es), with |
846 | way. AEMP relies on each node somehow knowing its own address(es) (e.g. by |
736 | convenience functionality. |
847 | configuration or DNS), and possibly the addresses of some seed nodes, but |
|
|
848 | will otherwise discover other nodes (and their IDs) itself. |
737 | |
849 | |
738 | This means that AEMP requires a less tightly controlled environment at the |
|
|
739 | cost of longer node references and a slightly higher management overhead. |
|
|
740 | |
|
|
741 | =item Erlang has a "remote ports are like local ports" philosophy, AEMP |
850 | =item * Erlang has a "remote ports are like local ports" philosophy, AEMP |
742 | uses "local ports are like remote ports". |
851 | uses "local ports are like remote ports". |
743 | |
852 | |
744 | The failure modes for local ports are quite different (runtime errors |
853 | The failure modes for local ports are quite different (runtime errors |
745 | only) then for remote ports - when a local port dies, you I<know> it dies, |
854 | only) then for remote ports - when a local port dies, you I<know> it dies, |
746 | when a connection to another node dies, you know nothing about the other |
855 | when a connection to another node dies, you know nothing about the other |
… | |
… | |
757 | |
866 | |
758 | Erlang uses processes that selectively receive messages, and therefore |
867 | Erlang uses processes that selectively receive messages, and therefore |
759 | needs a queue. AEMP is event based, queuing messages would serve no |
868 | needs a queue. AEMP is event based, queuing messages would serve no |
760 | useful purpose. For the same reason the pattern-matching abilities of |
869 | useful purpose. For the same reason the pattern-matching abilities of |
761 | AnyEvent::MP are more limited, as there is little need to be able to |
870 | AnyEvent::MP are more limited, as there is little need to be able to |
762 | filter messages without dequeing them. |
871 | filter messages without dequeuing them. |
763 | |
872 | |
764 | (But see L<Coro::MP> for a more Erlang-like process model on top of AEMP). |
873 | (But see L<Coro::MP> for a more Erlang-like process model on top of AEMP). |
765 | |
874 | |
766 | =item * Erlang sends are synchronous, AEMP sends are asynchronous. |
875 | =item * Erlang sends are synchronous, AEMP sends are asynchronous. |
767 | |
876 | |
… | |
… | |
769 | so does not need a queue that can overflow). AEMP sends are immediate, |
878 | so does not need a queue that can overflow). AEMP sends are immediate, |
770 | connection establishment is handled in the background. |
879 | connection establishment is handled in the background. |
771 | |
880 | |
772 | =item * Erlang suffers from silent message loss, AEMP does not. |
881 | =item * Erlang suffers from silent message loss, AEMP does not. |
773 | |
882 | |
774 | Erlang makes few guarantees on messages delivery - messages can get lost |
883 | Erlang implements few guarantees on messages delivery - messages can get |
775 | without any of the processes realising it (i.e. you send messages a, b, |
884 | lost without any of the processes realising it (i.e. you send messages a, |
776 | and c, and the other side only receives messages a and c). |
885 | b, and c, and the other side only receives messages a and c). |
777 | |
886 | |
778 | AEMP guarantees correct ordering, and the guarantee that there are no |
887 | AEMP guarantees correct ordering, and the guarantee that after one message |
779 | holes in the message sequence. |
888 | is lost, all following ones sent to the same port are lost as well, until |
780 | |
889 | monitoring raises an error, so there are no silent "holes" in the message |
781 | =item * In Erlang, processes can be declared dead and later be found to be |
890 | sequence. |
782 | alive. |
|
|
783 | |
|
|
784 | In Erlang it can happen that a monitored process is declared dead and |
|
|
785 | linked processes get killed, but later it turns out that the process is |
|
|
786 | still alive - and can receive messages. |
|
|
787 | |
|
|
788 | In AEMP, when port monitoring detects a port as dead, then that port will |
|
|
789 | eventually be killed - it cannot happen that a node detects a port as dead |
|
|
790 | and then later sends messages to it, finding it is still alive. |
|
|
791 | |
891 | |
792 | =item * Erlang can send messages to the wrong port, AEMP does not. |
892 | =item * Erlang can send messages to the wrong port, AEMP does not. |
793 | |
893 | |
794 | In Erlang it is quite likely that a node that restarts reuses a process ID |
894 | In Erlang it is quite likely that a node that restarts reuses a process ID |
795 | known to other nodes for a completely different process, causing messages |
895 | known to other nodes for a completely different process, causing messages |
… | |
… | |
799 | around in the network will not be sent to an unrelated port. |
899 | around in the network will not be sent to an unrelated port. |
800 | |
900 | |
801 | =item * Erlang uses unprotected connections, AEMP uses secure |
901 | =item * Erlang uses unprotected connections, AEMP uses secure |
802 | authentication and can use TLS. |
902 | authentication and can use TLS. |
803 | |
903 | |
804 | AEMP can use a proven protocol - SSL/TLS - to protect connections and |
904 | AEMP can use a proven protocol - TLS - to protect connections and |
805 | securely authenticate nodes. |
905 | securely authenticate nodes. |
806 | |
906 | |
807 | =item * The AEMP protocol is optimised for both text-based and binary |
907 | =item * The AEMP protocol is optimised for both text-based and binary |
808 | communications. |
908 | communications. |
809 | |
909 | |
810 | The AEMP protocol, unlike the Erlang protocol, supports both |
910 | The AEMP protocol, unlike the Erlang protocol, supports both programming |
811 | language-independent text-only protocols (good for debugging) and binary, |
911 | language independent text-only protocols (good for debugging) and binary, |
812 | language-specific serialisers (e.g. Storable). |
912 | language-specific serialisers (e.g. Storable). By default, unless TLS is |
|
|
913 | used, the protocol is actually completely text-based. |
813 | |
914 | |
814 | It has also been carefully designed to be implementable in other languages |
915 | It has also been carefully designed to be implementable in other languages |
815 | with a minimum of work while gracefully degrading fucntionality to make the |
916 | with a minimum of work while gracefully degrading functionality to make the |
816 | protocol simple. |
917 | protocol simple. |
817 | |
918 | |
818 | =item * AEMP has more flexible monitoring options than Erlang. |
919 | =item * AEMP has more flexible monitoring options than Erlang. |
819 | |
920 | |
820 | In Erlang, you can chose to receive I<all> exit signals as messages |
921 | In Erlang, you can chose to receive I<all> exit signals as messages |
… | |
… | |
823 | Erlang, as one can choose between automatic kill, exit message or callback |
924 | Erlang, as one can choose between automatic kill, exit message or callback |
824 | on a per-process basis. |
925 | on a per-process basis. |
825 | |
926 | |
826 | =item * Erlang tries to hide remote/local connections, AEMP does not. |
927 | =item * Erlang tries to hide remote/local connections, AEMP does not. |
827 | |
928 | |
828 | Monitoring in Erlang is not an indicator of process death/crashes, |
929 | Monitoring in Erlang is not an indicator of process death/crashes, in the |
829 | as linking is (except linking is unreliable in Erlang). |
930 | same way as linking is (except linking is unreliable in Erlang). |
830 | |
931 | |
831 | In AEMP, you don't "look up" registered port names or send to named ports |
932 | In AEMP, you don't "look up" registered port names or send to named ports |
832 | that might or might not be persistent. Instead, you normally spawn a port |
933 | that might or might not be persistent. Instead, you normally spawn a port |
833 | on the remote node. The init function monitors the you, and you monitor |
934 | on the remote node. The init function monitors you, and you monitor the |
834 | the remote port. Since both monitors are local to the node, they are much |
935 | remote port. Since both monitors are local to the node, they are much more |
835 | more reliable. |
936 | reliable (no need for C<spawn_link>). |
836 | |
937 | |
837 | This also saves round-trips and avoids sending messages to the wrong port |
938 | This also saves round-trips and avoids sending messages to the wrong port |
838 | (hard to do in Erlang). |
939 | (hard to do in Erlang). |
839 | |
940 | |
840 | =back |
941 | =back |
841 | |
942 | |
842 | =head1 RATIONALE |
943 | =head1 RATIONALE |
843 | |
944 | |
844 | =over 4 |
945 | =over 4 |
845 | |
946 | |
846 | =item Why strings for ports and noderefs, why not objects? |
947 | =item Why strings for port and node IDs, why not objects? |
847 | |
948 | |
848 | We considered "objects", but found that the actual number of methods |
949 | We considered "objects", but found that the actual number of methods |
849 | thatc an be called are very low. Since port IDs and noderefs travel over |
950 | that can be called are quite low. Since port and node IDs travel over |
850 | the network frequently, the serialising/deserialising would add lots of |
951 | the network frequently, the serialising/deserialising would add lots of |
851 | overhead, as well as having to keep a proxy object. |
952 | overhead, as well as having to keep a proxy object everywhere. |
852 | |
953 | |
853 | Strings can easily be printed, easily serialised etc. and need no special |
954 | Strings can easily be printed, easily serialised etc. and need no special |
854 | procedures to be "valid". |
955 | procedures to be "valid". |
855 | |
956 | |
856 | And a a miniport consists of a single closure stored in a global hash - it |
957 | And as a result, a port with just a default receiver consists of a single |
857 | can't become much cheaper. |
958 | closure stored in a global hash - it can't become much cheaper. |
858 | |
959 | |
859 | =item Why favour JSON, why not real serialising format such as Storable? |
960 | =item Why favour JSON, why not a real serialising format such as Storable? |
860 | |
961 | |
861 | In fact, any AnyEvent::MP node will happily accept Storable as framing |
962 | In fact, any AnyEvent::MP node will happily accept Storable as framing |
862 | format, but currently there is no way to make a node use Storable by |
963 | format, but currently there is no way to make a node use Storable by |
863 | default. |
964 | default (although all nodes will accept it). |
864 | |
965 | |
865 | The default framing protocol is JSON because a) JSON::XS is many times |
966 | The default framing protocol is JSON because a) JSON::XS is many times |
866 | faster for small messages and b) most importantly, after years of |
967 | faster for small messages and b) most importantly, after years of |
867 | experience we found that object serialisation is causing more problems |
968 | experience we found that object serialisation is causing more problems |
868 | than it gains: Just like function calls, objects simply do not travel |
969 | than it solves: Just like function calls, objects simply do not travel |
869 | easily over the network, mostly because they will always be a copy, so you |
970 | easily over the network, mostly because they will always be a copy, so you |
870 | always have to re-think your design. |
971 | always have to re-think your design. |
871 | |
972 | |
872 | Keeping your messages simple, concentrating on data structures rather than |
973 | Keeping your messages simple, concentrating on data structures rather than |
873 | objects, will keep your messages clean, tidy and efficient. |
974 | objects, will keep your messages clean, tidy and efficient. |
874 | |
975 | |
875 | =back |
976 | =back |
876 | |
977 | |
877 | =head1 SEE ALSO |
978 | =head1 SEE ALSO |
878 | |
979 | |
|
|
980 | L<AnyEvent::MP::Intro> - a gentle introduction. |
|
|
981 | |
|
|
982 | L<AnyEvent::MP::Kernel> - more, lower-level, stuff. |
|
|
983 | |
|
|
984 | L<AnyEvent::MP::Global> - network maintainance and port groups, to find |
|
|
985 | your applications. |
|
|
986 | |
|
|
987 | L<AnyEvent::MP::DataConn> - establish data connections between nodes. |
|
|
988 | |
|
|
989 | L<AnyEvent::MP::LogCatcher> - simple service to display log messages from |
|
|
990 | all nodes. |
|
|
991 | |
879 | L<AnyEvent>. |
992 | L<AnyEvent>. |
880 | |
993 | |
881 | =head1 AUTHOR |
994 | =head1 AUTHOR |
882 | |
995 | |
883 | Marc Lehmann <schmorp@schmorp.de> |
996 | Marc Lehmann <schmorp@schmorp.de> |