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 $simple_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 port 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 $port, $cb->(@msg) # callback is invoked on death |
40 | mon $port, $otherport # kill otherport on abnormal death |
41 | mon $port, $localport # kill localport on abnormal death |
41 | mon $port, $otherport, @msg # send message on death |
42 | mon $port, $localport, @msg # send message on death |
42 | |
43 | |
43 | =head1 CURRENT STATUS |
44 | # temporarily execute code in port context |
|
|
45 | peval $port, sub { die "kill the port!" }; |
44 | |
46 | |
45 | AnyEvent::MP - stable API, should work |
47 | # execute callbacks in $SELF port context |
46 | AnyEvent::MP::Intro - outdated |
48 | my $timer = AE::timer 1, 0, psub { |
47 | AnyEvent::MP::Kernel - WIP |
49 | die "kill the port, delayed"; |
48 | AnyEvent::MP::Transport - mostly stable |
50 | }; |
49 | |
51 | |
50 | stay tuned. |
52 | # distributed database - modification |
|
|
53 | db_set $family => $subkey [=> $value] # add a subkey |
|
|
54 | db_del $family => $subkey... # delete one or more subkeys |
|
|
55 | db_reg $family => $port [=> $value] # register a port |
|
|
56 | |
|
|
57 | # distributed database - queries |
|
|
58 | db_family $family => $cb->(\%familyhash) |
|
|
59 | db_keys $family => $cb->(\@keys) |
|
|
60 | db_values $family => $cb->(\@values) |
|
|
61 | |
|
|
62 | # distributed database - monitoring a family |
|
|
63 | db_mon $family => $cb->(\%familyhash, \@added, \@changed, \@deleted) |
51 | |
64 | |
52 | =head1 DESCRIPTION |
65 | =head1 DESCRIPTION |
53 | |
66 | |
54 | This module (-family) implements a simple message passing framework. |
67 | This module (-family) implements a simple message passing framework. |
55 | |
68 | |
56 | Despite its simplicity, you can securely message other processes running |
69 | Despite its simplicity, you can securely message other processes running |
57 | on the same or other hosts. |
70 | on the same or other hosts, and you can supervise entities remotely. |
58 | |
71 | |
59 | For an introduction to this module family, see the L<AnyEvent::MP::Intro> |
72 | For an introduction to this module family, see the L<AnyEvent::MP::Intro> |
60 | manual page. |
73 | 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 | |
74 | |
66 | =head1 CONCEPTS |
75 | =head1 CONCEPTS |
67 | |
76 | |
68 | =over 4 |
77 | =over 4 |
69 | |
78 | |
70 | =item port |
79 | =item port |
71 | |
80 | |
72 | A port is something you can send messages to (with the C<snd> function). |
81 | Not to be confused with a TCP port, a "port" is something you can send |
|
|
82 | messages to (with the C<snd> function). |
73 | |
83 | |
74 | Ports allow you to register C<rcv> handlers that can match all or just |
84 | Ports allow you to register C<rcv> handlers that can match all or just |
75 | some messages. Messages will not be queued. |
85 | some messages. Messages send to ports will not be queued, regardless of |
|
|
86 | anything was listening for them or not. |
76 | |
87 | |
|
|
88 | Ports are represented by (printable) strings called "port IDs". |
|
|
89 | |
77 | =item port id - C<noderef#portname> |
90 | =item port ID - C<nodeid#portname> |
78 | |
91 | |
79 | A port ID is the concatenation of a noderef, a hash-mark (C<#>) as |
92 | A port ID is the concatenation of a node ID, a hash-mark (C<#>) |
80 | separator, and a port name (a printable string of unspecified format). An |
93 | as separator, and a port name (a printable string of unspecified |
81 | exception is the the node port, whose ID is identical to its node |
94 | format created by AnyEvent::MP). |
82 | reference. |
|
|
83 | |
95 | |
84 | =item node |
96 | =item node |
85 | |
97 | |
86 | A node is a single process containing at least one port - the node port, |
98 | A node is a single process containing at least one port - the node port, |
87 | which provides nodes to manage each other remotely, and to create new |
99 | which enables nodes to manage each other remotely, and to create new |
88 | ports. |
100 | ports. |
89 | |
101 | |
90 | Nodes are either private (single-process only), slaves (connected to a |
102 | Nodes are either public (have one or more listening ports) or private |
91 | master node only) or public nodes (connectable from unrelated nodes). |
103 | (no listening ports). Private nodes cannot talk to other private nodes |
|
|
104 | currently, but all nodes can talk to public nodes. |
92 | |
105 | |
93 | =item noderef - C<host:port,host:port...>, C<id@noderef>, C<id> |
106 | Nodes is represented by (printable) strings called "node IDs". |
94 | |
107 | |
95 | A node reference is a string that either simply identifies the node (for |
108 | =item node ID - C<[A-Za-z0-9_\-.:]*> |
96 | private and slave nodes), or contains a recipe on how to reach a given |
|
|
97 | node (for public nodes). |
|
|
98 | |
109 | |
99 | This recipe is simply a comma-separated list of C<address:port> pairs (for |
110 | A node ID is a string that uniquely identifies the node within a |
100 | TCP/IP, other protocols might look different). |
111 | network. Depending on the configuration used, node IDs can look like a |
|
|
112 | hostname, a hostname and a port, or a random string. AnyEvent::MP itself |
|
|
113 | doesn't interpret node IDs in any way except to uniquely identify a node. |
101 | |
114 | |
102 | Node references come in two flavours: resolved (containing only numerical |
115 | =item binds - C<ip:port> |
103 | addresses) or unresolved (where hostnames are used instead of addresses). |
|
|
104 | |
116 | |
105 | Before using an unresolved node reference in a message you first have to |
117 | Nodes can only talk to each other by creating some kind of connection to |
106 | resolve it. |
118 | each other. To do this, nodes should listen on one or more local transport |
|
|
119 | endpoints - binds. |
|
|
120 | |
|
|
121 | Currently, only standard C<ip:port> specifications can be used, which |
|
|
122 | specify TCP ports to listen on. So a bind is basically just a tcp socket |
|
|
123 | in listening mode that accepts connections from other nodes. |
|
|
124 | |
|
|
125 | =item seed nodes |
|
|
126 | |
|
|
127 | When a node starts, it knows nothing about the network it is in - it |
|
|
128 | needs to connect to at least one other node that is already in the |
|
|
129 | network. These other nodes are called "seed nodes". |
|
|
130 | |
|
|
131 | Seed nodes themselves are not special - they are seed nodes only because |
|
|
132 | some other node I<uses> them as such, but any node can be used as seed |
|
|
133 | node for other nodes, and eahc node can use a different set of seed nodes. |
|
|
134 | |
|
|
135 | In addition to discovering the network, seed nodes are also used to |
|
|
136 | maintain the network - all nodes using the same seed node are part of the |
|
|
137 | same network. If a network is split into multiple subnets because e.g. the |
|
|
138 | network link between the parts goes down, then using the same seed nodes |
|
|
139 | for all nodes ensures that eventually the subnets get merged again. |
|
|
140 | |
|
|
141 | Seed nodes are expected to be long-running, and at least one seed node |
|
|
142 | should always be available. They should also be relatively responsive - a |
|
|
143 | seed node that blocks for long periods will slow down everybody else. |
|
|
144 | |
|
|
145 | For small networks, it's best if every node uses the same set of seed |
|
|
146 | nodes. For large networks, it can be useful to specify "regional" seed |
|
|
147 | nodes for most nodes in an area, and use all seed nodes as seed nodes for |
|
|
148 | each other. What's important is that all seed nodes connections form a |
|
|
149 | complete graph, so that the network cannot split into separate subnets |
|
|
150 | forever. |
|
|
151 | |
|
|
152 | Seed nodes are represented by seed IDs. |
|
|
153 | |
|
|
154 | =item seed IDs - C<host:port> |
|
|
155 | |
|
|
156 | Seed IDs are transport endpoint(s) (usually a hostname/IP address and a |
|
|
157 | TCP port) of nodes that should be used as seed nodes. |
|
|
158 | |
|
|
159 | =item global nodes |
|
|
160 | |
|
|
161 | An AEMP network needs a discovery service - nodes need to know how to |
|
|
162 | connect to other nodes they only know by name. In addition, AEMP offers a |
|
|
163 | distributed "group database", which maps group names to a list of strings |
|
|
164 | - for example, to register worker ports. |
|
|
165 | |
|
|
166 | A network needs at least one global node to work, and allows every node to |
|
|
167 | be a global node. |
|
|
168 | |
|
|
169 | Any node that loads the L<AnyEvent::MP::Global> module becomes a global |
|
|
170 | node and tries to keep connections to all other nodes. So while it can |
|
|
171 | make sense to make every node "global" in small networks, it usually makes |
|
|
172 | sense to only make seed nodes into global nodes in large networks (nodes |
|
|
173 | keep connections to seed nodes and global nodes, so making them the same |
|
|
174 | reduces overhead). |
107 | |
175 | |
108 | =back |
176 | =back |
109 | |
177 | |
110 | =head1 VARIABLES/FUNCTIONS |
178 | =head1 VARIABLES/FUNCTIONS |
111 | |
179 | |
… | |
… | |
113 | |
181 | |
114 | =cut |
182 | =cut |
115 | |
183 | |
116 | package AnyEvent::MP; |
184 | package AnyEvent::MP; |
117 | |
185 | |
|
|
186 | use AnyEvent::MP::Config (); |
118 | use AnyEvent::MP::Kernel; |
187 | use AnyEvent::MP::Kernel; |
|
|
188 | use AnyEvent::MP::Kernel qw( |
|
|
189 | %NODE %PORT %PORT_DATA $UNIQ $RUNIQ $ID |
|
|
190 | add_node load_func |
|
|
191 | |
|
|
192 | NODE $NODE |
|
|
193 | configure |
|
|
194 | node_of port_is_local |
|
|
195 | snd kil |
|
|
196 | db_set db_del |
|
|
197 | db_mon db_family db_keys db_values |
|
|
198 | ); |
119 | |
199 | |
120 | use common::sense; |
200 | use common::sense; |
121 | |
201 | |
122 | use Carp (); |
202 | use Carp (); |
123 | |
203 | |
124 | use AE (); |
204 | use AnyEvent (); |
|
|
205 | use Guard (); |
125 | |
206 | |
126 | use base "Exporter"; |
207 | use base "Exporter"; |
127 | |
208 | |
128 | our $VERSION = $AnyEvent::MP::Kernel::VERSION; |
209 | our $VERSION = '2.02'; # also in MP/Config.pm |
129 | |
210 | |
130 | our @EXPORT = qw( |
211 | our @EXPORT = qw( |
131 | NODE $NODE *SELF node_of after |
212 | configure |
132 | resolve_node initialise_node |
213 | |
133 | snd rcv mon mon_guard kil reg psub spawn |
214 | NODE $NODE |
134 | port |
215 | *SELF |
|
|
216 | |
|
|
217 | node_of port_is_local |
|
|
218 | |
|
|
219 | snd kil |
|
|
220 | port rcv mon mon_guard psub peval spawn cal |
|
|
221 | db_set db_del db_reg |
|
|
222 | db_mon db_family db_keys db_values |
|
|
223 | |
|
|
224 | after |
135 | ); |
225 | ); |
136 | |
226 | |
137 | our $SELF; |
227 | our $SELF; |
138 | |
228 | |
139 | sub _self_die() { |
229 | sub _self_die() { |
… | |
… | |
142 | kil $SELF, die => $msg; |
232 | kil $SELF, die => $msg; |
143 | } |
233 | } |
144 | |
234 | |
145 | =item $thisnode = NODE / $NODE |
235 | =item $thisnode = NODE / $NODE |
146 | |
236 | |
147 | The C<NODE> function returns, and the C<$NODE> variable contains the |
237 | The C<NODE> function returns, and the C<$NODE> variable contains, the node |
148 | noderef of the local node. The value is initialised by a call to |
238 | ID of the node running in the current process. This value is initialised by |
149 | C<initialise_node>. |
239 | a call to C<configure>. |
150 | |
240 | |
151 | =item $noderef = node_of $port |
241 | =item $nodeid = node_of $port |
152 | |
242 | |
153 | Extracts and returns the noderef from a port ID or a noderef. |
243 | Extracts and returns the node ID from a port ID or a node ID. |
154 | |
244 | |
155 | =item initialise_node $noderef, $seednode, $seednode... |
245 | =item $is_local = port_is_local $port |
156 | |
246 | |
157 | =item initialise_node "slave/", $master, $master... |
247 | Returns true iff the port is a local port. |
158 | |
248 | |
|
|
249 | =item configure $profile, key => value... |
|
|
250 | |
|
|
251 | =item configure key => value... |
|
|
252 | |
159 | Before a node can talk to other nodes on the network it has to initialise |
253 | Before a node can talk to other nodes on the network (i.e. enter |
160 | itself - the minimum a node needs to know is it's own name, and optionally |
254 | "distributed mode") it has to configure itself - the minimum a node needs |
161 | it should know the noderefs of some other nodes in the network. |
255 | to know is its own name, and optionally it should know the addresses of |
|
|
256 | some other nodes in the network to discover other nodes. |
162 | |
257 | |
163 | This function initialises a node - it must be called exactly once (or |
258 | This function configures a node - it must be called exactly once (or |
164 | never) before calling other AnyEvent::MP functions. |
259 | never) before calling other AnyEvent::MP functions. |
165 | |
260 | |
166 | All arguments (optionally except for the first) are noderefs, which can be |
261 | The key/value pairs are basically the same ones as documented for the |
167 | either resolved or unresolved. |
262 | F<aemp> command line utility (sans the set/del prefix), with these additions: |
168 | |
|
|
169 | The first argument will be looked up in the configuration database first |
|
|
170 | (if it is C<undef> then the current nodename will be used instead) to find |
|
|
171 | the relevant configuration profile (see L<aemp>). If none is found then |
|
|
172 | the default configuration is used. The configuration supplies additional |
|
|
173 | seed/master nodes and can override the actual noderef. |
|
|
174 | |
|
|
175 | There are two types of networked nodes, public nodes and slave nodes: |
|
|
176 | |
263 | |
177 | =over 4 |
264 | =over 4 |
178 | |
265 | |
179 | =item public nodes |
266 | =item norc => $boolean (default false) |
180 | |
267 | |
181 | For public nodes, C<$noderef> (supplied either directly to |
268 | If true, then the rc file (e.g. F<~/.perl-anyevent-mp>) will I<not> |
182 | C<initialise_node> or indirectly via a profile or the nodename) must be a |
269 | be consulted - all configuration options must be specified in the |
183 | noderef (possibly unresolved, in which case it will be resolved). |
270 | C<configure> call. |
184 | |
271 | |
185 | After resolving, the node will bind itself on all endpoints and try to |
272 | =item force => $boolean (default false) |
186 | connect to all additional C<$seednodes> that are specified. Seednodes are |
|
|
187 | optional and can be used to quickly bootstrap the node into an existing |
|
|
188 | network. |
|
|
189 | |
273 | |
190 | =item slave nodes |
274 | IF true, then the values specified in the C<configure> will take |
191 | |
275 | precedence over any values configured via the rc file. The default is for |
192 | When the C<$noderef> (either as given or overriden by the config file) |
276 | the rc file to override any options specified in the program. |
193 | is the special string C<slave/>, then the node will become a slave |
|
|
194 | node. Slave nodes cannot be contacted from outside and will route most of |
|
|
195 | their traffic to the master node that they attach to. |
|
|
196 | |
|
|
197 | At least one additional noderef is required (either by specifying it |
|
|
198 | directly or because it is part of the configuration profile): The node |
|
|
199 | will try to connect to all of them and will become a slave attached to the |
|
|
200 | first node it can successfully connect to. |
|
|
201 | |
|
|
202 | Note that slave nodes cannot change their name, and consequently, their |
|
|
203 | master, so if the master goes down, the slave node will not function well |
|
|
204 | anymore until it can re-establish conenciton to its master. This makes |
|
|
205 | slave nodes unsuitable for long-term nodes or fault-tolerant networks. |
|
|
206 | |
277 | |
207 | =back |
278 | =back |
208 | |
279 | |
209 | This function will block until all nodes have been resolved and, for slave |
|
|
210 | nodes, until it has successfully established a connection to a master |
|
|
211 | server. |
|
|
212 | |
|
|
213 | All the seednodes will also be specially marked to automatically retry |
|
|
214 | connecting to them infinitely. |
|
|
215 | |
|
|
216 | Example: become a public node listening on the guessed noderef, or the one |
|
|
217 | specified via C<aemp> for the current node. This should be the most common |
|
|
218 | form of invocation for "daemon"-type nodes. |
|
|
219 | |
|
|
220 | initialise_node; |
|
|
221 | |
|
|
222 | Example: become a slave node to any of the the seednodes specified via |
|
|
223 | C<aemp>. This form is often used for commandline clients. |
|
|
224 | |
|
|
225 | initialise_node "slave/"; |
|
|
226 | |
|
|
227 | Example: become a slave node to any of the specified master servers. This |
|
|
228 | form is also often used for commandline clients. |
|
|
229 | |
|
|
230 | initialise_node "slave/", "master1", "192.168.13.17", "mp.example.net"; |
|
|
231 | |
|
|
232 | Example: become a public node, and try to contact some well-known master |
|
|
233 | servers to become part of the network. |
|
|
234 | |
|
|
235 | initialise_node undef, "master1", "master2"; |
|
|
236 | |
|
|
237 | Example: become a public node listening on port C<4041>. |
|
|
238 | |
|
|
239 | initialise_node 4041; |
|
|
240 | |
|
|
241 | Example: become a public node, only visible on localhost port 4044. |
|
|
242 | |
|
|
243 | initialise_node "localhost:4044"; |
|
|
244 | |
|
|
245 | =item $cv = resolve_node $noderef |
|
|
246 | |
|
|
247 | Takes an unresolved node reference that may contain hostnames and |
|
|
248 | abbreviated IDs, resolves all of them and returns a resolved node |
|
|
249 | reference. |
|
|
250 | |
|
|
251 | In addition to C<address:port> pairs allowed in resolved noderefs, the |
|
|
252 | following forms are supported: |
|
|
253 | |
|
|
254 | =over 4 |
280 | =over 4 |
255 | |
281 | |
256 | =item the empty string |
282 | =item step 1, gathering configuration from profiles |
257 | |
283 | |
258 | An empty-string component gets resolved as if the default port (4040) was |
284 | The function first looks up a profile in the aemp configuration (see the |
259 | specified. |
285 | L<aemp> commandline utility). The profile name can be specified via the |
|
|
286 | named C<profile> parameter or can simply be the first parameter). If it is |
|
|
287 | missing, then the nodename (F<uname -n>) will be used as profile name. |
260 | |
288 | |
261 | =item naked port numbers (e.g. C<1234>) |
289 | The profile data is then gathered as follows: |
262 | |
290 | |
263 | These are resolved by prepending the local nodename and a colon, to be |
291 | First, all remaining key => value pairs (all of which are conveniently |
264 | further resolved. |
292 | undocumented at the moment) will be interpreted as configuration |
|
|
293 | data. Then they will be overwritten by any values specified in the global |
|
|
294 | default configuration (see the F<aemp> utility), then the chain of |
|
|
295 | profiles chosen by the profile name (and any C<parent> attributes). |
265 | |
296 | |
266 | =item hostnames (e.g. C<localhost:1234>, C<localhost>) |
297 | That means that the values specified in the profile have highest priority |
|
|
298 | and the values specified directly via C<configure> have lowest priority, |
|
|
299 | and can only be used to specify defaults. |
267 | |
300 | |
268 | These are resolved by using AnyEvent::DNS to resolve them, optionally |
301 | If the profile specifies a node ID, then this will become the node ID of |
269 | looking up SRV records for the C<aemp=4040> port, if no port was |
302 | this process. If not, then the profile name will be used as node ID, with |
270 | specified. |
303 | a unique randoms tring (C</%u>) appended. |
|
|
304 | |
|
|
305 | The node ID can contain some C<%> sequences that are expanded: C<%n> |
|
|
306 | is expanded to the local nodename, C<%u> is replaced by a random |
|
|
307 | strign to make the node unique. For example, the F<aemp> commandline |
|
|
308 | utility uses C<aemp/%n/%u> as nodename, which might expand to |
|
|
309 | C<aemp/cerebro/ZQDGSIkRhEZQDGSIkRhE>. |
|
|
310 | |
|
|
311 | =item step 2, bind listener sockets |
|
|
312 | |
|
|
313 | The next step is to look up the binds in the profile, followed by binding |
|
|
314 | aemp protocol listeners on all binds specified (it is possible and valid |
|
|
315 | to have no binds, meaning that the node cannot be contacted from the |
|
|
316 | outside. This means the node cannot talk to other nodes that also have no |
|
|
317 | binds, but it can still talk to all "normal" nodes). |
|
|
318 | |
|
|
319 | If the profile does not specify a binds list, then a default of C<*> is |
|
|
320 | used, meaning the node will bind on a dynamically-assigned port on every |
|
|
321 | local IP address it finds. |
|
|
322 | |
|
|
323 | =item step 3, connect to seed nodes |
|
|
324 | |
|
|
325 | As the last step, the seed ID list from the profile is passed to the |
|
|
326 | L<AnyEvent::MP::Global> module, which will then use it to keep |
|
|
327 | connectivity with at least one node at any point in time. |
271 | |
328 | |
272 | =back |
329 | =back |
|
|
330 | |
|
|
331 | Example: become a distributed node using the local node name as profile. |
|
|
332 | This should be the most common form of invocation for "daemon"-type nodes. |
|
|
333 | |
|
|
334 | configure |
|
|
335 | |
|
|
336 | Example: become a semi-anonymous node. This form is often used for |
|
|
337 | commandline clients. |
|
|
338 | |
|
|
339 | configure nodeid => "myscript/%n/%u"; |
|
|
340 | |
|
|
341 | Example: configure a node using a profile called seed, which is suitable |
|
|
342 | for a seed node as it binds on all local addresses on a fixed port (4040, |
|
|
343 | customary for aemp). |
|
|
344 | |
|
|
345 | # use the aemp commandline utility |
|
|
346 | # aemp profile seed binds '*:4040' |
|
|
347 | |
|
|
348 | # then use it |
|
|
349 | configure profile => "seed"; |
|
|
350 | |
|
|
351 | # or simply use aemp from the shell again: |
|
|
352 | # aemp run profile seed |
|
|
353 | |
|
|
354 | # or provide a nicer-to-remember nodeid |
|
|
355 | # aemp run profile seed nodeid "$(hostname)" |
273 | |
356 | |
274 | =item $SELF |
357 | =item $SELF |
275 | |
358 | |
276 | Contains the current port id while executing C<rcv> callbacks or C<psub> |
359 | Contains the current port id while executing C<rcv> callbacks or C<psub> |
277 | blocks. |
360 | blocks. |
278 | |
361 | |
279 | =item SELF, %SELF, @SELF... |
362 | =item *SELF, SELF, %SELF, @SELF... |
280 | |
363 | |
281 | Due to some quirks in how perl exports variables, it is impossible to |
364 | Due to some quirks in how perl exports variables, it is impossible to |
282 | just export C<$SELF>, all the symbols called C<SELF> are exported by this |
365 | just export C<$SELF>, all the symbols named C<SELF> are exported by this |
283 | module, but only C<$SELF> is currently used. |
366 | module, but only C<$SELF> is currently used. |
284 | |
367 | |
285 | =item snd $port, type => @data |
368 | =item snd $port, type => @data |
286 | |
369 | |
287 | =item snd $port, @msg |
370 | =item snd $port, @msg |
288 | |
371 | |
289 | Send the given message to the given port ID, which can identify either |
372 | Send the given message to the given port, which can identify either a |
290 | a local or a remote port, and must be a port ID. |
373 | local or a remote port, and must be a port ID. |
291 | |
374 | |
292 | While the message can be about anything, it is highly recommended to use a |
375 | While the message can be almost anything, it is highly recommended to |
293 | string as first element (a port ID, or some word that indicates a request |
376 | use a string as first element (a port ID, or some word that indicates a |
294 | type etc.). |
377 | request type etc.) and to consist if only simple perl values (scalars, |
|
|
378 | arrays, hashes) - if you think you need to pass an object, think again. |
295 | |
379 | |
296 | The message data effectively becomes read-only after a call to this |
380 | The message data logically becomes read-only after a call to this |
297 | function: modifying any argument is not allowed and can cause many |
381 | function: modifying any argument (or values referenced by them) is |
298 | problems. |
382 | forbidden, as there can be considerable time between the call to C<snd> |
|
|
383 | and the time the message is actually being serialised - in fact, it might |
|
|
384 | never be copied as within the same process it is simply handed to the |
|
|
385 | receiving port. |
299 | |
386 | |
300 | The type of data you can transfer depends on the transport protocol: when |
387 | The type of data you can transfer depends on the transport protocol: when |
301 | JSON is used, then only strings, numbers and arrays and hashes consisting |
388 | JSON is used, then only strings, numbers and arrays and hashes consisting |
302 | of those are allowed (no objects). When Storable is used, then anything |
389 | of those are allowed (no objects). When Storable is used, then anything |
303 | that Storable can serialise and deserialise is allowed, and for the local |
390 | that Storable can serialise and deserialise is allowed, and for the local |
304 | node, anything can be passed. |
391 | node, anything can be passed. Best rely only on the common denominator of |
|
|
392 | these. |
305 | |
393 | |
306 | =item $local_port = port |
394 | =item $local_port = port |
307 | |
395 | |
308 | Create a new local port object and returns its port ID. Initially it has |
396 | Create a new local port object and returns its port ID. Initially it has |
309 | no callbacks set and will throw an error when it receives messages. |
397 | no callbacks set and will throw an error when it receives messages. |
… | |
… | |
328 | |
416 | |
329 | =cut |
417 | =cut |
330 | |
418 | |
331 | sub rcv($@); |
419 | sub rcv($@); |
332 | |
420 | |
333 | sub _kilme { |
421 | my $KILME = sub { |
334 | die "received message on port without callback"; |
422 | (my $tag = substr $_[0], 0, 30) =~ s/([^\x20-\x7e])/./g; |
335 | } |
423 | kil $SELF, unhandled_message => "no callback found for message '$tag'"; |
|
|
424 | }; |
336 | |
425 | |
337 | sub port(;&) { |
426 | sub port(;&) { |
338 | my $id = "$UNIQ." . $ID++; |
427 | my $id = $UNIQ . ++$ID; |
339 | my $port = "$NODE#$id"; |
428 | my $port = "$NODE#$id"; |
340 | |
429 | |
341 | rcv $port, shift || \&_kilme; |
430 | rcv $port, shift || $KILME; |
342 | |
431 | |
343 | $port |
432 | $port |
344 | } |
433 | } |
345 | |
434 | |
346 | =item rcv $local_port, $callback->(@msg) |
435 | =item rcv $local_port, $callback->(@msg) |
… | |
… | |
351 | |
440 | |
352 | The global C<$SELF> (exported by this module) contains C<$port> while |
441 | The global C<$SELF> (exported by this module) contains C<$port> while |
353 | executing the callback. Runtime errors during callback execution will |
442 | executing the callback. Runtime errors during callback execution will |
354 | result in the port being C<kil>ed. |
443 | result in the port being C<kil>ed. |
355 | |
444 | |
356 | The default callback received all messages not matched by a more specific |
445 | The default callback receives all messages not matched by a more specific |
357 | C<tag> match. |
446 | C<tag> match. |
358 | |
447 | |
359 | =item rcv $local_port, tag => $callback->(@msg_without_tag), ... |
448 | =item rcv $local_port, tag => $callback->(@msg_without_tag), ... |
360 | |
449 | |
361 | Register (or replace) callbacks to be called on messages starting with the |
450 | Register (or replace) callbacks to be called on messages starting with the |
… | |
… | |
382 | msg1 => sub { ... }, |
471 | msg1 => sub { ... }, |
383 | ... |
472 | ... |
384 | ; |
473 | ; |
385 | |
474 | |
386 | Example: temporarily register a rcv callback for a tag matching some port |
475 | Example: temporarily register a rcv callback for a tag matching some port |
387 | (e.g. for a rpc reply) and unregister it after a message was received. |
476 | (e.g. for an rpc reply) and unregister it after a message was received. |
388 | |
477 | |
389 | rcv $port, $otherport => sub { |
478 | rcv $port, $otherport => sub { |
390 | my @reply = @_; |
479 | my @reply = @_; |
391 | |
480 | |
392 | rcv $SELF, $otherport; |
481 | rcv $SELF, $otherport; |
… | |
… | |
394 | |
483 | |
395 | =cut |
484 | =cut |
396 | |
485 | |
397 | sub rcv($@) { |
486 | sub rcv($@) { |
398 | my $port = shift; |
487 | my $port = shift; |
399 | my ($noderef, $portid) = split /#/, $port, 2; |
488 | my ($nodeid, $portid) = split /#/, $port, 2; |
400 | |
489 | |
401 | $NODE{$noderef} == $NODE{""} |
490 | $nodeid eq $NODE |
402 | or Carp::croak "$port: rcv can only be called on local ports, caught"; |
491 | or Carp::croak "$port: rcv can only be called on local ports, caught"; |
403 | |
492 | |
404 | while (@_) { |
493 | while (@_) { |
405 | if (ref $_[0]) { |
494 | if (ref $_[0]) { |
406 | if (my $self = $PORT_DATA{$portid}) { |
495 | if (my $self = $PORT_DATA{$portid}) { |
407 | "AnyEvent::MP::Port" eq ref $self |
496 | "AnyEvent::MP::Port" eq ref $self |
408 | or Carp::croak "$port: rcv can only be called on message matching ports, caught"; |
497 | or Carp::croak "$port: rcv can only be called on message matching ports, caught"; |
409 | |
498 | |
410 | $self->[2] = shift; |
499 | $self->[0] = shift; |
411 | } else { |
500 | } else { |
412 | my $cb = shift; |
501 | my $cb = shift; |
413 | $PORT{$portid} = sub { |
502 | $PORT{$portid} = sub { |
414 | local $SELF = $port; |
503 | local $SELF = $port; |
415 | eval { &$cb }; _self_die if $@; |
504 | eval { &$cb }; _self_die if $@; |
416 | }; |
505 | }; |
417 | } |
506 | } |
418 | } elsif (defined $_[0]) { |
507 | } elsif (defined $_[0]) { |
419 | my $self = $PORT_DATA{$portid} ||= do { |
508 | my $self = $PORT_DATA{$portid} ||= do { |
420 | my $self = bless [$PORT{$port} || sub { }, { }, $port], "AnyEvent::MP::Port"; |
509 | my $self = bless [$PORT{$portid} || sub { }, { }, $port], "AnyEvent::MP::Port"; |
421 | |
510 | |
422 | $PORT{$portid} = sub { |
511 | $PORT{$portid} = sub { |
423 | local $SELF = $port; |
512 | local $SELF = $port; |
424 | |
513 | |
425 | if (my $cb = $self->[1]{$_[0]}) { |
514 | if (my $cb = $self->[1]{$_[0]}) { |
… | |
… | |
447 | } |
536 | } |
448 | |
537 | |
449 | $port |
538 | $port |
450 | } |
539 | } |
451 | |
540 | |
|
|
541 | =item peval $port, $coderef[, @args] |
|
|
542 | |
|
|
543 | Evaluates the given C<$codref> within the context of C<$port>, that is, |
|
|
544 | when the code throws an exception the C<$port> will be killed. |
|
|
545 | |
|
|
546 | Any remaining args will be passed to the callback. Any return values will |
|
|
547 | be returned to the caller. |
|
|
548 | |
|
|
549 | This is useful when you temporarily want to execute code in the context of |
|
|
550 | a port. |
|
|
551 | |
|
|
552 | Example: create a port and run some initialisation code in it's context. |
|
|
553 | |
|
|
554 | my $port = port { ... }; |
|
|
555 | |
|
|
556 | peval $port, sub { |
|
|
557 | init |
|
|
558 | or die "unable to init"; |
|
|
559 | }; |
|
|
560 | |
|
|
561 | =cut |
|
|
562 | |
|
|
563 | sub peval($$) { |
|
|
564 | local $SELF = shift; |
|
|
565 | my $cb = shift; |
|
|
566 | |
|
|
567 | if (wantarray) { |
|
|
568 | my @res = eval { &$cb }; |
|
|
569 | _self_die if $@; |
|
|
570 | @res |
|
|
571 | } else { |
|
|
572 | my $res = eval { &$cb }; |
|
|
573 | _self_die if $@; |
|
|
574 | $res |
|
|
575 | } |
|
|
576 | } |
|
|
577 | |
452 | =item $closure = psub { BLOCK } |
578 | =item $closure = psub { BLOCK } |
453 | |
579 | |
454 | Remembers C<$SELF> and creates a closure out of the BLOCK. When the |
580 | Remembers C<$SELF> and creates a closure out of the BLOCK. When the |
455 | closure is executed, sets up the environment in the same way as in C<rcv> |
581 | closure is executed, sets up the environment in the same way as in C<rcv> |
456 | callbacks, i.e. runtime errors will cause the port to get C<kil>ed. |
582 | callbacks, i.e. runtime errors will cause the port to get C<kil>ed. |
|
|
583 | |
|
|
584 | The effect is basically as if it returned C<< sub { peval $SELF, sub { |
|
|
585 | BLOCK }, @_ } >>. |
457 | |
586 | |
458 | This is useful when you register callbacks from C<rcv> callbacks: |
587 | This is useful when you register callbacks from C<rcv> callbacks: |
459 | |
588 | |
460 | rcv delayed_reply => sub { |
589 | rcv delayed_reply => sub { |
461 | my ($delay, @reply) = @_; |
590 | my ($delay, @reply) = @_; |
… | |
… | |
485 | $res |
614 | $res |
486 | } |
615 | } |
487 | } |
616 | } |
488 | } |
617 | } |
489 | |
618 | |
490 | =item $guard = mon $port, $cb->(@reason) |
619 | =item $guard = mon $port, $rcvport # kill $rcvport when $port dies |
491 | |
620 | |
492 | =item $guard = mon $port, $rcvport |
621 | =item $guard = mon $port # kill $SELF when $port dies |
493 | |
622 | |
494 | =item $guard = mon $port |
623 | =item $guard = mon $port, $cb->(@reason) # call $cb when $port dies |
495 | |
624 | |
496 | =item $guard = mon $port, $rcvport, @msg |
625 | =item $guard = mon $port, $rcvport, @msg # send a message when $port dies |
497 | |
626 | |
498 | Monitor the given port and do something when the port is killed or |
627 | Monitor the given port and do something when the port is killed or |
499 | messages to it were lost, and optionally return a guard that can be used |
628 | messages to it were lost, and optionally return a guard that can be used |
500 | to stop monitoring again. |
629 | to stop monitoring again. |
501 | |
630 | |
|
|
631 | The first two forms distinguish between "normal" and "abnormal" kil's: |
|
|
632 | |
|
|
633 | In the first form (another port given), if the C<$port> is C<kil>'ed with |
|
|
634 | a non-empty reason, the other port (C<$rcvport>) will be kil'ed with the |
|
|
635 | same reason. That is, on "normal" kil's nothing happens, while under all |
|
|
636 | other conditions, the other port is killed with the same reason. |
|
|
637 | |
|
|
638 | The second form (kill self) is the same as the first form, except that |
|
|
639 | C<$rvport> defaults to C<$SELF>. |
|
|
640 | |
|
|
641 | The remaining forms don't distinguish between "normal" and "abnormal" kil's |
|
|
642 | - it's up to the callback or receiver to check whether the C<@reason> is |
|
|
643 | empty and act accordingly. |
|
|
644 | |
|
|
645 | In the third form (callback), the callback is simply called with any |
|
|
646 | number of C<@reason> elements (empty @reason means that the port was deleted |
|
|
647 | "normally"). Note also that I<< the callback B<must> never die >>, so use |
|
|
648 | C<eval> if unsure. |
|
|
649 | |
|
|
650 | In the last form (message), a message of the form C<$rcvport, @msg, |
|
|
651 | @reason> will be C<snd>. |
|
|
652 | |
|
|
653 | Monitoring-actions are one-shot: once messages are lost (and a monitoring |
|
|
654 | alert was raised), they are removed and will not trigger again, even if it |
|
|
655 | turns out that the port is still alive (but monitoring actions added after |
|
|
656 | that will again trigger). |
|
|
657 | |
|
|
658 | As a rule of thumb, monitoring requests should always monitor a remote |
|
|
659 | port locally (using a local C<$rcvport> or a callback). The reason is that |
|
|
660 | kill messages might get lost, just like any other message. Another less |
|
|
661 | obvious reason is that even monitoring requests can get lost (for example, |
|
|
662 | when the connection to the other node goes down permanently). When |
|
|
663 | monitoring a port locally these problems do not exist. |
|
|
664 | |
502 | C<mon> effectively guarantees that, in the absence of hardware failures, |
665 | C<mon> effectively guarantees that, in the absence of hardware failures, |
503 | that after starting the monitor, either all messages sent to the port |
666 | after starting the monitor, either all messages sent to the port will |
504 | will arrive, or the monitoring action will be invoked after possible |
667 | arrive, or the monitoring action will be invoked after possible message |
505 | message loss has been detected. No messages will be lost "in between" |
668 | loss has been detected. No messages will be lost "in between" (after |
506 | (after the first lost message no further messages will be received by the |
669 | the first lost message no further messages will be received by the |
507 | port). After the monitoring action was invoked, further messages might get |
670 | port). After the monitoring action was invoked, further messages might get |
508 | delivered again. |
671 | delivered again. |
509 | |
672 | |
510 | Note that monitoring-actions are one-shot: once released, they are removed |
673 | Inter-host-connection timeouts and monitoring depend on the transport |
511 | and will not trigger again. |
674 | used. The only transport currently implemented is TCP, and AnyEvent::MP |
|
|
675 | relies on TCP to detect node-downs (this can take 10-15 minutes on a |
|
|
676 | non-idle connection, and usually around two hours for idle connections). |
512 | |
677 | |
513 | In the first form (callback), the callback is simply called with any |
678 | This means that monitoring is good for program errors and cleaning up |
514 | number of C<@reason> elements (no @reason means that the port was deleted |
679 | stuff eventually, but they are no replacement for a timeout when you need |
515 | "normally"). Note also that I<< the callback B<must> never die >>, so use |
680 | to ensure some maximum latency. |
516 | C<eval> if unsure. |
|
|
517 | |
|
|
518 | In the second form (another port given), the other port (C<$rcvport>) |
|
|
519 | will be C<kil>'ed with C<@reason>, iff a @reason was specified, i.e. on |
|
|
520 | "normal" kils nothing happens, while under all other conditions, the other |
|
|
521 | port is killed with the same reason. |
|
|
522 | |
|
|
523 | The third form (kill self) is the same as the second form, except that |
|
|
524 | C<$rvport> defaults to C<$SELF>. |
|
|
525 | |
|
|
526 | In the last form (message), a message of the form C<@msg, @reason> will be |
|
|
527 | C<snd>. |
|
|
528 | |
|
|
529 | As a rule of thumb, monitoring requests should always monitor a port from |
|
|
530 | a local port (or callback). The reason is that kill messages might get |
|
|
531 | lost, just like any other message. Another less obvious reason is that |
|
|
532 | even monitoring requests can get lost (for exmaple, when the connection |
|
|
533 | to the other node goes down permanently). When monitoring a port locally |
|
|
534 | these problems do not exist. |
|
|
535 | |
681 | |
536 | Example: call a given callback when C<$port> is killed. |
682 | Example: call a given callback when C<$port> is killed. |
537 | |
683 | |
538 | mon $port, sub { warn "port died because of <@_>\n" }; |
684 | mon $port, sub { warn "port died because of <@_>\n" }; |
539 | |
685 | |
… | |
… | |
546 | mon $port, $self => "restart"; |
692 | mon $port, $self => "restart"; |
547 | |
693 | |
548 | =cut |
694 | =cut |
549 | |
695 | |
550 | sub mon { |
696 | sub mon { |
551 | my ($noderef, $port) = split /#/, shift, 2; |
697 | my ($nodeid, $port) = split /#/, shift, 2; |
552 | |
698 | |
553 | my $node = $NODE{$noderef} || add_node $noderef; |
699 | my $node = $NODE{$nodeid} || add_node $nodeid; |
554 | |
700 | |
555 | my $cb = @_ ? shift : $SELF || Carp::croak 'mon: called with one argument only, but $SELF not set,'; |
701 | my $cb = @_ ? shift : $SELF || Carp::croak 'mon: called with one argument only, but $SELF not set,'; |
556 | |
702 | |
557 | unless (ref $cb) { |
703 | unless (ref $cb) { |
558 | if (@_) { |
704 | if (@_) { |
… | |
… | |
567 | } |
713 | } |
568 | |
714 | |
569 | $node->monitor ($port, $cb); |
715 | $node->monitor ($port, $cb); |
570 | |
716 | |
571 | defined wantarray |
717 | defined wantarray |
572 | and AnyEvent::Util::guard { $node->unmonitor ($port, $cb) } |
718 | and ($cb += 0, Guard::guard { $node->unmonitor ($port, $cb) }) |
573 | } |
719 | } |
574 | |
720 | |
575 | =item $guard = mon_guard $port, $ref, $ref... |
721 | =item $guard = mon_guard $port, $ref, $ref... |
576 | |
722 | |
577 | Monitors the given C<$port> and keeps the passed references. When the port |
723 | Monitors the given C<$port> and keeps the passed references. When the port |
578 | is killed, the references will be freed. |
724 | is killed, the references will be freed. |
579 | |
725 | |
580 | Optionally returns a guard that will stop the monitoring. |
726 | Optionally returns a guard that will stop the monitoring. |
581 | |
727 | |
582 | This function is useful when you create e.g. timers or other watchers and |
728 | This function is useful when you create e.g. timers or other watchers and |
583 | want to free them when the port gets killed: |
729 | want to free them when the port gets killed (note the use of C<psub>): |
584 | |
730 | |
585 | $port->rcv (start => sub { |
731 | $port->rcv (start => sub { |
586 | my $timer; $timer = mon_guard $port, AE::timer 1, 1, sub { |
732 | my $timer; $timer = mon_guard $port, AE::timer 1, 1, psub { |
587 | undef $timer if 0.9 < rand; |
733 | undef $timer if 0.9 < rand; |
588 | }); |
734 | }); |
589 | }); |
735 | }); |
590 | |
736 | |
591 | =cut |
737 | =cut |
… | |
… | |
600 | |
746 | |
601 | =item kil $port[, @reason] |
747 | =item kil $port[, @reason] |
602 | |
748 | |
603 | Kill the specified port with the given C<@reason>. |
749 | Kill the specified port with the given C<@reason>. |
604 | |
750 | |
605 | If no C<@reason> is specified, then the port is killed "normally" (linked |
751 | If no C<@reason> is specified, then the port is killed "normally" - |
606 | ports will not be kileld, or even notified). |
752 | monitor callback will be invoked, but the kil will not cause linked ports |
|
|
753 | (C<mon $mport, $lport> form) to get killed. |
607 | |
754 | |
608 | Otherwise, linked ports get killed with the same reason (second form of |
755 | If a C<@reason> is specified, then linked ports (C<mon $mport, $lport> |
609 | C<mon>, see below). |
756 | form) get killed with the same reason. |
610 | |
757 | |
611 | Runtime errors while evaluating C<rcv> callbacks or inside C<psub> blocks |
758 | Runtime errors while evaluating C<rcv> callbacks or inside C<psub> blocks |
612 | will be reported as reason C<< die => $@ >>. |
759 | will be reported as reason C<< die => $@ >>. |
613 | |
760 | |
614 | Transport/communication errors are reported as C<< transport_error => |
761 | Transport/communication errors are reported as C<< transport_error => |
615 | $message >>. |
762 | $message >>. |
616 | |
763 | |
617 | =cut |
764 | Common idioms: |
|
|
765 | |
|
|
766 | # silently remove yourself, do not kill linked ports |
|
|
767 | kil $SELF; |
|
|
768 | |
|
|
769 | # report a failure in some detail |
|
|
770 | kil $SELF, failure_mode_1 => "it failed with too high temperature"; |
|
|
771 | |
|
|
772 | # do not waste much time with killing, just die when something goes wrong |
|
|
773 | open my $fh, "<file" |
|
|
774 | or die "file: $!"; |
618 | |
775 | |
619 | =item $port = spawn $node, $initfunc[, @initdata] |
776 | =item $port = spawn $node, $initfunc[, @initdata] |
620 | |
777 | |
621 | Creates a port on the node C<$node> (which can also be a port ID, in which |
778 | Creates a port on the node C<$node> (which can also be a port ID, in which |
622 | case it's the node where that port resides). |
779 | case it's the node where that port resides). |
623 | |
780 | |
624 | The port ID of the newly created port is return immediately, and it is |
781 | The port ID of the newly created port is returned immediately, and it is |
625 | permissible to immediately start sending messages or monitor the port. |
782 | possible to immediately start sending messages or to monitor the port. |
626 | |
783 | |
627 | After the port has been created, the init function is |
784 | After the port has been created, the init function is called on the remote |
628 | called. This function must be a fully-qualified function name |
785 | node, in the same context as a C<rcv> callback. This function must be a |
629 | (e.g. C<MyApp::Chat::Server::init>). To specify a function in the main |
786 | fully-qualified function name (e.g. C<MyApp::Chat::Server::init>). To |
630 | program, use C<::name>. |
787 | specify a function in the main program, use C<::name>. |
631 | |
788 | |
632 | If the function doesn't exist, then the node tries to C<require> |
789 | If the function doesn't exist, then the node tries to C<require> |
633 | the package, then the package above the package and so on (e.g. |
790 | the package, then the package above the package and so on (e.g. |
634 | C<MyApp::Chat::Server>, C<MyApp::Chat>, C<MyApp>) until the function |
791 | C<MyApp::Chat::Server>, C<MyApp::Chat>, C<MyApp>) until the function |
635 | exists or it runs out of package names. |
792 | exists or it runs out of package names. |
636 | |
793 | |
637 | The init function is then called with the newly-created port as context |
794 | The init function is then called with the newly-created port as context |
638 | object (C<$SELF>) and the C<@initdata> values as arguments. |
795 | object (C<$SELF>) and the C<@initdata> values as arguments. It I<must> |
|
|
796 | call one of the C<rcv> functions to set callbacks on C<$SELF>, otherwise |
|
|
797 | the port might not get created. |
639 | |
798 | |
640 | A common idiom is to pass your own port, monitor the spawned port, and |
799 | A common idiom is to pass a local port, immediately monitor the spawned |
641 | in the init function, monitor the original port. This two-way monitoring |
800 | port, and in the remote init function, immediately monitor the passed |
642 | ensures that both ports get cleaned up when there is a problem. |
801 | local port. This two-way monitoring ensures that both ports get cleaned up |
|
|
802 | when there is a problem. |
|
|
803 | |
|
|
804 | C<spawn> guarantees that the C<$initfunc> has no visible effects on the |
|
|
805 | caller before C<spawn> returns (by delaying invocation when spawn is |
|
|
806 | called for the local node). |
643 | |
807 | |
644 | Example: spawn a chat server port on C<$othernode>. |
808 | Example: spawn a chat server port on C<$othernode>. |
645 | |
809 | |
646 | # this node, executed from within a port context: |
810 | # this node, executed from within a port context: |
647 | my $server = spawn $othernode, "MyApp::Chat::Server::connect", $SELF; |
811 | my $server = spawn $othernode, "MyApp::Chat::Server::connect", $SELF; |
… | |
… | |
662 | |
826 | |
663 | sub _spawn { |
827 | sub _spawn { |
664 | my $port = shift; |
828 | my $port = shift; |
665 | my $init = shift; |
829 | my $init = shift; |
666 | |
830 | |
|
|
831 | # rcv will create the actual port |
667 | local $SELF = "$NODE#$port"; |
832 | local $SELF = "$NODE#$port"; |
668 | eval { |
833 | eval { |
669 | &{ load_func $init } |
834 | &{ load_func $init } |
670 | }; |
835 | }; |
671 | _self_die if $@; |
836 | _self_die if $@; |
672 | } |
837 | } |
673 | |
838 | |
674 | sub spawn(@) { |
839 | sub spawn(@) { |
675 | my ($noderef, undef) = split /#/, shift, 2; |
840 | my ($nodeid, undef) = split /#/, shift, 2; |
676 | |
841 | |
677 | my $id = "$RUNIQ." . $ID++; |
842 | my $id = $RUNIQ . ++$ID; |
678 | |
843 | |
679 | $_[0] =~ /::/ |
844 | $_[0] =~ /::/ |
680 | or Carp::croak "spawn init function must be a fully-qualified name, caught"; |
845 | or Carp::croak "spawn init function must be a fully-qualified name, caught"; |
681 | |
846 | |
682 | snd_to_func $noderef, "AnyEvent::MP::_spawn" => $id, @_; |
847 | snd_to_func $nodeid, "AnyEvent::MP::_spawn" => $id, @_; |
683 | |
848 | |
684 | "$noderef#$id" |
849 | "$nodeid#$id" |
685 | } |
850 | } |
|
|
851 | |
686 | |
852 | |
687 | =item after $timeout, @msg |
853 | =item after $timeout, @msg |
688 | |
854 | |
689 | =item after $timeout, $callback |
855 | =item after $timeout, $callback |
690 | |
856 | |
691 | Either sends the given message, or call the given callback, after the |
857 | Either sends the given message, or call the given callback, after the |
692 | specified number of seconds. |
858 | specified number of seconds. |
693 | |
859 | |
694 | This is simply a utility function that come sin handy at times. |
860 | This is simply a utility function that comes in handy at times - the |
|
|
861 | AnyEvent::MP author is not convinced of the wisdom of having it, though, |
|
|
862 | so it may go away in the future. |
695 | |
863 | |
696 | =cut |
864 | =cut |
697 | |
865 | |
698 | sub after($@) { |
866 | sub after($@) { |
699 | my ($timeout, @action) = @_; |
867 | my ($timeout, @action) = @_; |
… | |
… | |
704 | ? $action[0]() |
872 | ? $action[0]() |
705 | : snd @action; |
873 | : snd @action; |
706 | }; |
874 | }; |
707 | } |
875 | } |
708 | |
876 | |
|
|
877 | #=item $cb2 = timeout $seconds, $cb[, @args] |
|
|
878 | |
|
|
879 | =item cal $port, @msg, $callback[, $timeout] |
|
|
880 | |
|
|
881 | A simple form of RPC - sends a message to the given C<$port> with the |
|
|
882 | given contents (C<@msg>), but appends a reply port to the message. |
|
|
883 | |
|
|
884 | The reply port is created temporarily just for the purpose of receiving |
|
|
885 | the reply, and will be C<kil>ed when no longer needed. |
|
|
886 | |
|
|
887 | A reply message sent to the port is passed to the C<$callback> as-is. |
|
|
888 | |
|
|
889 | If an optional time-out (in seconds) is given and it is not C<undef>, |
|
|
890 | then the callback will be called without any arguments after the time-out |
|
|
891 | elapsed and the port is C<kil>ed. |
|
|
892 | |
|
|
893 | If no time-out is given (or it is C<undef>), then the local port will |
|
|
894 | monitor the remote port instead, so it eventually gets cleaned-up. |
|
|
895 | |
|
|
896 | Currently this function returns the temporary port, but this "feature" |
|
|
897 | might go in future versions unless you can make a convincing case that |
|
|
898 | this is indeed useful for something. |
|
|
899 | |
|
|
900 | =cut |
|
|
901 | |
|
|
902 | sub cal(@) { |
|
|
903 | my $timeout = ref $_[-1] ? undef : pop; |
|
|
904 | my $cb = pop; |
|
|
905 | |
|
|
906 | my $port = port { |
|
|
907 | undef $timeout; |
|
|
908 | kil $SELF; |
|
|
909 | &$cb; |
|
|
910 | }; |
|
|
911 | |
|
|
912 | if (defined $timeout) { |
|
|
913 | $timeout = AE::timer $timeout, 0, sub { |
|
|
914 | undef $timeout; |
|
|
915 | kil $port; |
|
|
916 | $cb->(); |
|
|
917 | }; |
|
|
918 | } else { |
|
|
919 | mon $_[0], sub { |
|
|
920 | kil $port; |
|
|
921 | $cb->(); |
|
|
922 | }; |
|
|
923 | } |
|
|
924 | |
|
|
925 | push @_, $port; |
|
|
926 | &snd; |
|
|
927 | |
|
|
928 | $port |
|
|
929 | } |
|
|
930 | |
|
|
931 | =back |
|
|
932 | |
|
|
933 | =head1 DISTRIBUTED DATABASE |
|
|
934 | |
|
|
935 | AnyEvent::MP comes with a simple distributed database. The database will |
|
|
936 | be mirrored asynchronously on all global nodes. Other nodes bind to one |
|
|
937 | of the global nodes for their needs. Every node has a "local database" |
|
|
938 | which contains all the values that are set locally. All local databases |
|
|
939 | are merged together to form the global database, which can be queried. |
|
|
940 | |
|
|
941 | The database structure is that of a two-level hash - the database hash |
|
|
942 | contains hashes which contain values, similarly to a perl hash of hashes, |
|
|
943 | i.e.: |
|
|
944 | |
|
|
945 | $DATABASE{$family}{$subkey} = $value |
|
|
946 | |
|
|
947 | The top level hash key is called "family", and the second-level hash key |
|
|
948 | is called "subkey" or simply "key". |
|
|
949 | |
|
|
950 | The family must be alphanumeric, i.e. start with a letter and consist |
|
|
951 | of letters, digits, underscores and colons (C<[A-Za-z][A-Za-z0-9_:]*>, |
|
|
952 | pretty much like Perl module names. |
|
|
953 | |
|
|
954 | As the family namespace is global, it is recommended to prefix family names |
|
|
955 | with the name of the application or module using it. |
|
|
956 | |
|
|
957 | The subkeys must be non-empty strings, with no further restrictions. |
|
|
958 | |
|
|
959 | The values should preferably be strings, but other perl scalars should |
|
|
960 | work as well (such as C<undef>, arrays and hashes). |
|
|
961 | |
|
|
962 | Every database entry is owned by one node - adding the same family/subkey |
|
|
963 | combination on multiple nodes will not cause discomfort for AnyEvent::MP, |
|
|
964 | but the result might be nondeterministic, i.e. the key might have |
|
|
965 | different values on different nodes. |
|
|
966 | |
|
|
967 | Different subkeys in the same family can be owned by different nodes |
|
|
968 | without problems, and in fact, this is the common method to create worker |
|
|
969 | pools. For example, a worker port for image scaling might do this: |
|
|
970 | |
|
|
971 | db_set my_image_scalers => $port; |
|
|
972 | |
|
|
973 | And clients looking for an image scaler will want to get the |
|
|
974 | C<my_image_scalers> keys from time to time: |
|
|
975 | |
|
|
976 | db_keys my_image_scalers => sub { |
|
|
977 | @ports = @{ $_[0] }; |
|
|
978 | }; |
|
|
979 | |
|
|
980 | Or better yet, they want to monitor the database family, so they always |
|
|
981 | have a reasonable up-to-date copy: |
|
|
982 | |
|
|
983 | db_mon my_image_scalers => sub { |
|
|
984 | @ports = keys %{ $_[0] }; |
|
|
985 | }; |
|
|
986 | |
|
|
987 | In general, you can set or delete single subkeys, but query and monitor |
|
|
988 | whole families only. |
|
|
989 | |
|
|
990 | If you feel the need to monitor or query a single subkey, try giving it |
|
|
991 | it's own family. |
|
|
992 | |
|
|
993 | =over |
|
|
994 | |
|
|
995 | =item $guard = db_set $family => $subkey [=> $value] |
|
|
996 | |
|
|
997 | Sets (or replaces) a key to the database - if C<$value> is omitted, |
|
|
998 | C<undef> is used instead. |
|
|
999 | |
|
|
1000 | When called in non-void context, C<db_set> returns a guard that |
|
|
1001 | automatically calls C<db_del> when it is destroyed. |
|
|
1002 | |
|
|
1003 | =item db_del $family => $subkey... |
|
|
1004 | |
|
|
1005 | Deletes one or more subkeys from the database family. |
|
|
1006 | |
|
|
1007 | =item $guard = db_reg $family => $port => $value |
|
|
1008 | |
|
|
1009 | =item $guard = db_reg $family => $port |
|
|
1010 | |
|
|
1011 | =item $guard = db_reg $family |
|
|
1012 | |
|
|
1013 | Registers a port in the given family and optionally returns a guard to |
|
|
1014 | remove it. |
|
|
1015 | |
|
|
1016 | This function basically does the same as: |
|
|
1017 | |
|
|
1018 | db_set $family => $port => $value |
|
|
1019 | |
|
|
1020 | Except that the port is monitored and automatically removed from the |
|
|
1021 | database family when it is kil'ed. |
|
|
1022 | |
|
|
1023 | If C<$value> is missing, C<undef> is used. If C<$port> is missing, then |
|
|
1024 | C<$SELF> is used. |
|
|
1025 | |
|
|
1026 | This function is most useful to register a port in some port group (which |
|
|
1027 | is just another name for a database family), and have it removed when the |
|
|
1028 | port is gone. This works best when the port is a local port. |
|
|
1029 | |
|
|
1030 | =cut |
|
|
1031 | |
|
|
1032 | sub db_reg($$;$) { |
|
|
1033 | my $family = shift; |
|
|
1034 | my $port = @_ ? shift : $SELF; |
|
|
1035 | |
|
|
1036 | my $clr = sub { db_del $family => $port }; |
|
|
1037 | mon $port, $clr; |
|
|
1038 | |
|
|
1039 | db_set $family => $port => $_[0]; |
|
|
1040 | |
|
|
1041 | defined wantarray |
|
|
1042 | and &Guard::guard ($clr) |
|
|
1043 | } |
|
|
1044 | |
|
|
1045 | =item db_family $family => $cb->(\%familyhash) |
|
|
1046 | |
|
|
1047 | Queries the named database C<$family> and call the callback with the |
|
|
1048 | family represented as a hash. You can keep and freely modify the hash. |
|
|
1049 | |
|
|
1050 | =item db_keys $family => $cb->(\@keys) |
|
|
1051 | |
|
|
1052 | Same as C<db_family>, except it only queries the family I<subkeys> and passes |
|
|
1053 | them as array reference to the callback. |
|
|
1054 | |
|
|
1055 | =item db_values $family => $cb->(\@values) |
|
|
1056 | |
|
|
1057 | Same as C<db_family>, except it only queries the family I<values> and passes them |
|
|
1058 | as array reference to the callback. |
|
|
1059 | |
|
|
1060 | =item $guard = db_mon $family => $cb->(\%familyhash, \@added, \@changed, \@deleted) |
|
|
1061 | |
|
|
1062 | Creates a monitor on the given database family. Each time a key is |
|
|
1063 | set or is deleted the callback is called with a hash containing the |
|
|
1064 | database family and three lists of added, changed and deleted subkeys, |
|
|
1065 | respectively. If no keys have changed then the array reference might be |
|
|
1066 | C<undef> or even missing. |
|
|
1067 | |
|
|
1068 | If not called in void context, a guard object is returned that, when |
|
|
1069 | destroyed, stops the monitor. |
|
|
1070 | |
|
|
1071 | The family hash reference and the key arrays belong to AnyEvent::MP and |
|
|
1072 | B<must not be modified or stored> by the callback. When in doubt, make a |
|
|
1073 | copy. |
|
|
1074 | |
|
|
1075 | As soon as possible after the monitoring starts, the callback will be |
|
|
1076 | called with the intiial contents of the family, even if it is empty, |
|
|
1077 | i.e. there will always be a timely call to the callback with the current |
|
|
1078 | contents. |
|
|
1079 | |
|
|
1080 | It is possible that the callback is called with a change event even though |
|
|
1081 | the subkey is already present and the value has not changed. |
|
|
1082 | |
|
|
1083 | The monitoring stops when the guard object is destroyed. |
|
|
1084 | |
|
|
1085 | Example: on every change to the family "mygroup", print out all keys. |
|
|
1086 | |
|
|
1087 | my $guard = db_mon mygroup => sub { |
|
|
1088 | my ($family, $a, $c, $d) = @_; |
|
|
1089 | print "mygroup members: ", (join " ", keys %$family), "\n"; |
|
|
1090 | }; |
|
|
1091 | |
|
|
1092 | Exmaple: wait until the family "My::Module::workers" is non-empty. |
|
|
1093 | |
|
|
1094 | my $guard; $guard = db_mon My::Module::workers => sub { |
|
|
1095 | my ($family, $a, $c, $d) = @_; |
|
|
1096 | return unless %$family; |
|
|
1097 | undef $guard; |
|
|
1098 | print "My::Module::workers now nonempty\n"; |
|
|
1099 | }; |
|
|
1100 | |
|
|
1101 | Example: print all changes to the family "AnyEvent::Fantasy::Module". |
|
|
1102 | |
|
|
1103 | my $guard = db_mon AnyEvent::Fantasy::Module => sub { |
|
|
1104 | my ($family, $a, $c, $d) = @_; |
|
|
1105 | |
|
|
1106 | print "+$_=$family->{$_}\n" for @$a; |
|
|
1107 | print "*$_=$family->{$_}\n" for @$c; |
|
|
1108 | print "-$_=$family->{$_}\n" for @$d; |
|
|
1109 | }; |
|
|
1110 | |
|
|
1111 | =cut |
|
|
1112 | |
709 | =back |
1113 | =back |
710 | |
1114 | |
711 | =head1 AnyEvent::MP vs. Distributed Erlang |
1115 | =head1 AnyEvent::MP vs. Distributed Erlang |
712 | |
1116 | |
713 | AnyEvent::MP got lots of its ideas from distributed Erlang (Erlang node |
1117 | AnyEvent::MP got lots of its ideas from distributed Erlang (Erlang node |
714 | == aemp node, Erlang process == aemp port), so many of the documents and |
1118 | == aemp node, Erlang process == aemp port), so many of the documents and |
715 | programming techniques employed by Erlang apply to AnyEvent::MP. Here is a |
1119 | programming techniques employed by Erlang apply to AnyEvent::MP. Here is a |
716 | sample: |
1120 | sample: |
717 | |
1121 | |
718 | http://www.Erlang.se/doc/programming_rules.shtml |
1122 | http://www.erlang.se/doc/programming_rules.shtml |
719 | http://Erlang.org/doc/getting_started/part_frame.html # chapters 3 and 4 |
1123 | http://erlang.org/doc/getting_started/part_frame.html # chapters 3 and 4 |
720 | http://Erlang.org/download/Erlang-book-part1.pdf # chapters 5 and 6 |
1124 | http://erlang.org/download/erlang-book-part1.pdf # chapters 5 and 6 |
721 | http://Erlang.org/download/armstrong_thesis_2003.pdf # chapters 4 and 5 |
1125 | http://erlang.org/download/armstrong_thesis_2003.pdf # chapters 4 and 5 |
722 | |
1126 | |
723 | Despite the similarities, there are also some important differences: |
1127 | Despite the similarities, there are also some important differences: |
724 | |
1128 | |
725 | =over 4 |
1129 | =over 4 |
726 | |
1130 | |
727 | =item * Node references contain the recipe on how to contact them. |
1131 | =item * Node IDs are arbitrary strings in AEMP. |
728 | |
1132 | |
729 | Erlang relies on special naming and DNS to work everywhere in the |
1133 | Erlang relies on special naming and DNS to work everywhere in the same |
730 | same way. AEMP relies on each node knowing it's own address(es), with |
1134 | way. AEMP relies on each node somehow knowing its own address(es) (e.g. by |
731 | convenience functionality. |
1135 | configuration or DNS), and possibly the addresses of some seed nodes, but |
732 | |
1136 | will otherwise discover other nodes (and their IDs) itself. |
733 | This means that AEMP requires a less tightly controlled environment at the |
|
|
734 | cost of longer node references and a slightly higher management overhead. |
|
|
735 | |
1137 | |
736 | =item * Erlang has a "remote ports are like local ports" philosophy, AEMP |
1138 | =item * Erlang has a "remote ports are like local ports" philosophy, AEMP |
737 | uses "local ports are like remote ports". |
1139 | uses "local ports are like remote ports". |
738 | |
1140 | |
739 | The failure modes for local ports are quite different (runtime errors |
1141 | The failure modes for local ports are quite different (runtime errors |
… | |
… | |
748 | ports being the special case/exception, where transport errors cannot |
1150 | ports being the special case/exception, where transport errors cannot |
749 | occur. |
1151 | occur. |
750 | |
1152 | |
751 | =item * Erlang uses processes and a mailbox, AEMP does not queue. |
1153 | =item * Erlang uses processes and a mailbox, AEMP does not queue. |
752 | |
1154 | |
753 | Erlang uses processes that selectively receive messages, and therefore |
1155 | Erlang uses processes that selectively receive messages out of order, and |
754 | needs a queue. AEMP is event based, queuing messages would serve no |
1156 | therefore needs a queue. AEMP is event based, queuing messages would serve |
755 | useful purpose. For the same reason the pattern-matching abilities of |
1157 | no useful purpose. For the same reason the pattern-matching abilities |
756 | AnyEvent::MP are more limited, as there is little need to be able to |
1158 | of AnyEvent::MP are more limited, as there is little need to be able to |
757 | filter messages without dequeing them. |
1159 | filter messages without dequeuing them. |
758 | |
1160 | |
759 | (But see L<Coro::MP> for a more Erlang-like process model on top of AEMP). |
1161 | This is not a philosophical difference, but simply stems from AnyEvent::MP |
|
|
1162 | being event-based, while Erlang is process-based. |
|
|
1163 | |
|
|
1164 | You can have a look at L<Coro::MP> for a more Erlang-like process model on |
|
|
1165 | top of AEMP and Coro threads. |
760 | |
1166 | |
761 | =item * Erlang sends are synchronous, AEMP sends are asynchronous. |
1167 | =item * Erlang sends are synchronous, AEMP sends are asynchronous. |
762 | |
1168 | |
763 | Sending messages in Erlang is synchronous and blocks the process (and |
1169 | Sending messages in Erlang is synchronous and blocks the process until |
|
|
1170 | a connection has been established and the message sent (and so does not |
764 | so does not need a queue that can overflow). AEMP sends are immediate, |
1171 | need a queue that can overflow). AEMP sends return immediately, connection |
765 | connection establishment is handled in the background. |
1172 | establishment is handled in the background. |
766 | |
1173 | |
767 | =item * Erlang suffers from silent message loss, AEMP does not. |
1174 | =item * Erlang suffers from silent message loss, AEMP does not. |
768 | |
1175 | |
769 | Erlang makes few guarantees on messages delivery - messages can get lost |
1176 | Erlang implements few guarantees on messages delivery - messages can get |
770 | without any of the processes realising it (i.e. you send messages a, b, |
1177 | lost without any of the processes realising it (i.e. you send messages a, |
771 | and c, and the other side only receives messages a and c). |
1178 | b, and c, and the other side only receives messages a and c). |
772 | |
1179 | |
773 | AEMP guarantees correct ordering, and the guarantee that there are no |
1180 | AEMP guarantees (modulo hardware errors) correct ordering, and the |
|
|
1181 | guarantee that after one message is lost, all following ones sent to the |
|
|
1182 | same port are lost as well, until monitoring raises an error, so there are |
774 | holes in the message sequence. |
1183 | no silent "holes" in the message sequence. |
775 | |
1184 | |
776 | =item * In Erlang, processes can be declared dead and later be found to be |
1185 | If you want your software to be very reliable, you have to cope with |
777 | alive. |
1186 | corrupted and even out-of-order messages in both Erlang and AEMP. AEMP |
778 | |
1187 | simply tries to work better in common error cases, such as when a network |
779 | In Erlang it can happen that a monitored process is declared dead and |
1188 | link goes down. |
780 | linked processes get killed, but later it turns out that the process is |
|
|
781 | still alive - and can receive messages. |
|
|
782 | |
|
|
783 | In AEMP, when port monitoring detects a port as dead, then that port will |
|
|
784 | eventually be killed - it cannot happen that a node detects a port as dead |
|
|
785 | and then later sends messages to it, finding it is still alive. |
|
|
786 | |
1189 | |
787 | =item * Erlang can send messages to the wrong port, AEMP does not. |
1190 | =item * Erlang can send messages to the wrong port, AEMP does not. |
788 | |
1191 | |
789 | In Erlang it is quite likely that a node that restarts reuses a process ID |
1192 | In Erlang it is quite likely that a node that restarts reuses an Erlang |
790 | known to other nodes for a completely different process, causing messages |
1193 | process ID known to other nodes for a completely different process, |
791 | destined for that process to end up in an unrelated process. |
1194 | causing messages destined for that process to end up in an unrelated |
|
|
1195 | process. |
792 | |
1196 | |
793 | AEMP never reuses port IDs, so old messages or old port IDs floating |
1197 | AEMP does not reuse port IDs, so old messages or old port IDs floating |
794 | around in the network will not be sent to an unrelated port. |
1198 | around in the network will not be sent to an unrelated port. |
795 | |
1199 | |
796 | =item * Erlang uses unprotected connections, AEMP uses secure |
1200 | =item * Erlang uses unprotected connections, AEMP uses secure |
797 | authentication and can use TLS. |
1201 | authentication and can use TLS. |
798 | |
1202 | |
799 | AEMP can use a proven protocol - SSL/TLS - to protect connections and |
1203 | AEMP can use a proven protocol - TLS - to protect connections and |
800 | securely authenticate nodes. |
1204 | securely authenticate nodes. |
801 | |
1205 | |
802 | =item * The AEMP protocol is optimised for both text-based and binary |
1206 | =item * The AEMP protocol is optimised for both text-based and binary |
803 | communications. |
1207 | communications. |
804 | |
1208 | |
805 | The AEMP protocol, unlike the Erlang protocol, supports both |
1209 | The AEMP protocol, unlike the Erlang protocol, supports both programming |
806 | language-independent text-only protocols (good for debugging) and binary, |
1210 | language independent text-only protocols (good for debugging), and binary, |
807 | language-specific serialisers (e.g. Storable). |
1211 | language-specific serialisers (e.g. Storable). By default, unless TLS is |
|
|
1212 | used, the protocol is actually completely text-based. |
808 | |
1213 | |
809 | It has also been carefully designed to be implementable in other languages |
1214 | It has also been carefully designed to be implementable in other languages |
810 | with a minimum of work while gracefully degrading fucntionality to make the |
1215 | with a minimum of work while gracefully degrading functionality to make the |
811 | protocol simple. |
1216 | protocol simple. |
812 | |
1217 | |
813 | =item * AEMP has more flexible monitoring options than Erlang. |
1218 | =item * AEMP has more flexible monitoring options than Erlang. |
814 | |
1219 | |
815 | In Erlang, you can chose to receive I<all> exit signals as messages |
1220 | In Erlang, you can chose to receive I<all> exit signals as messages or |
816 | or I<none>, there is no in-between, so monitoring single processes is |
1221 | I<none>, there is no in-between, so monitoring single Erlang processes is |
817 | difficult to implement. Monitoring in AEMP is more flexible than in |
1222 | difficult to implement. |
818 | Erlang, as one can choose between automatic kill, exit message or callback |
1223 | |
819 | on a per-process basis. |
1224 | Monitoring in AEMP is more flexible than in Erlang, as one can choose |
|
|
1225 | between automatic kill, exit message or callback on a per-port basis. |
820 | |
1226 | |
821 | =item * Erlang tries to hide remote/local connections, AEMP does not. |
1227 | =item * Erlang tries to hide remote/local connections, AEMP does not. |
822 | |
1228 | |
823 | Monitoring in Erlang is not an indicator of process death/crashes, |
1229 | Monitoring in Erlang is not an indicator of process death/crashes, in the |
824 | as linking is (except linking is unreliable in Erlang). |
1230 | same way as linking is (except linking is unreliable in Erlang). |
825 | |
1231 | |
826 | In AEMP, you don't "look up" registered port names or send to named ports |
1232 | In AEMP, you don't "look up" registered port names or send to named ports |
827 | that might or might not be persistent. Instead, you normally spawn a port |
1233 | that might or might not be persistent. Instead, you normally spawn a port |
828 | on the remote node. The init function monitors the you, and you monitor |
1234 | on the remote node. The init function monitors you, and you monitor the |
829 | the remote port. Since both monitors are local to the node, they are much |
1235 | remote port. Since both monitors are local to the node, they are much more |
830 | more reliable. |
1236 | reliable (no need for C<spawn_link>). |
831 | |
1237 | |
832 | This also saves round-trips and avoids sending messages to the wrong port |
1238 | This also saves round-trips and avoids sending messages to the wrong port |
833 | (hard to do in Erlang). |
1239 | (hard to do in Erlang). |
834 | |
1240 | |
835 | =back |
1241 | =back |
836 | |
1242 | |
837 | =head1 RATIONALE |
1243 | =head1 RATIONALE |
838 | |
1244 | |
839 | =over 4 |
1245 | =over 4 |
840 | |
1246 | |
841 | =item Why strings for ports and noderefs, why not objects? |
1247 | =item Why strings for port and node IDs, why not objects? |
842 | |
1248 | |
843 | We considered "objects", but found that the actual number of methods |
1249 | We considered "objects", but found that the actual number of methods |
844 | thatc an be called are very low. Since port IDs and noderefs travel over |
1250 | that can be called are quite low. Since port and node IDs travel over |
845 | the network frequently, the serialising/deserialising would add lots of |
1251 | the network frequently, the serialising/deserialising would add lots of |
846 | overhead, as well as having to keep a proxy object. |
1252 | overhead, as well as having to keep a proxy object everywhere. |
847 | |
1253 | |
848 | Strings can easily be printed, easily serialised etc. and need no special |
1254 | Strings can easily be printed, easily serialised etc. and need no special |
849 | procedures to be "valid". |
1255 | procedures to be "valid". |
850 | |
1256 | |
851 | And a a miniport consists of a single closure stored in a global hash - it |
1257 | And as a result, a port with just a default receiver consists of a single |
852 | can't become much cheaper. |
1258 | code reference stored in a global hash - it can't become much cheaper. |
853 | |
1259 | |
854 | =item Why favour JSON, why not real serialising format such as Storable? |
1260 | =item Why favour JSON, why not a real serialising format such as Storable? |
855 | |
1261 | |
856 | In fact, any AnyEvent::MP node will happily accept Storable as framing |
1262 | In fact, any AnyEvent::MP node will happily accept Storable as framing |
857 | format, but currently there is no way to make a node use Storable by |
1263 | format, but currently there is no way to make a node use Storable by |
858 | default. |
1264 | default (although all nodes will accept it). |
859 | |
1265 | |
860 | The default framing protocol is JSON because a) JSON::XS is many times |
1266 | The default framing protocol is JSON because a) JSON::XS is many times |
861 | faster for small messages and b) most importantly, after years of |
1267 | faster for small messages and b) most importantly, after years of |
862 | experience we found that object serialisation is causing more problems |
1268 | experience we found that object serialisation is causing more problems |
863 | than it gains: Just like function calls, objects simply do not travel |
1269 | than it solves: Just like function calls, objects simply do not travel |
864 | easily over the network, mostly because they will always be a copy, so you |
1270 | easily over the network, mostly because they will always be a copy, so you |
865 | always have to re-think your design. |
1271 | always have to re-think your design. |
866 | |
1272 | |
867 | Keeping your messages simple, concentrating on data structures rather than |
1273 | Keeping your messages simple, concentrating on data structures rather than |
868 | objects, will keep your messages clean, tidy and efficient. |
1274 | objects, will keep your messages clean, tidy and efficient. |
869 | |
1275 | |
870 | =back |
1276 | =back |
871 | |
1277 | |
|
|
1278 | =head1 PORTING FROM AnyEvent::MP VERSION 1.X |
|
|
1279 | |
|
|
1280 | AEMP version 2 has a few major incompatible changes compared to version 1: |
|
|
1281 | |
|
|
1282 | =over 4 |
|
|
1283 | |
|
|
1284 | =item AnyEvent::MP::Global no longer has group management functions. |
|
|
1285 | |
|
|
1286 | At least not officially - the grp_* functions are still exported and might |
|
|
1287 | work, but they will be removed in some later release. |
|
|
1288 | |
|
|
1289 | AnyEvent::MP now comes with a distributed database that is more |
|
|
1290 | powerful. Its database families map closely to port groups, but the API |
|
|
1291 | has changed (the functions are also now exported by AnyEvent::MP). Here is |
|
|
1292 | a rough porting guide: |
|
|
1293 | |
|
|
1294 | grp_reg $group, $port # old |
|
|
1295 | db_reg $group, $port # new |
|
|
1296 | |
|
|
1297 | $list = grp_get $group # old |
|
|
1298 | db_keys $group, sub { my $list = shift } # new |
|
|
1299 | |
|
|
1300 | grp_mon $group, $cb->(\@ports, $add, $del) # old |
|
|
1301 | db_mon $group, $cb->(\%ports, $add, $change, $del) # new |
|
|
1302 | |
|
|
1303 | C<grp_reg> is a no-brainer (just replace by C<db_reg>), but C<grp_get> is |
|
|
1304 | no longer instant, because the local node might not have a copy of the |
|
|
1305 | group. You can either modify your code to allow for a callback, or use |
|
|
1306 | C<db_mon> to keep an updated copy of the group: |
|
|
1307 | |
|
|
1308 | my $local_group_copy; |
|
|
1309 | db_mon $group => sub { $local_group_copy = $_[0] }; |
|
|
1310 | |
|
|
1311 | # now "keys %$local_group_copy" always returns the most up-to-date |
|
|
1312 | # list of ports in the group. |
|
|
1313 | |
|
|
1314 | C<grp_mon> can be replaced by C<db_mon> with minor changes - C<db_mon> |
|
|
1315 | passes a hash as first argument, and an extra C<$chg> argument that can be |
|
|
1316 | ignored: |
|
|
1317 | |
|
|
1318 | db_mon $group => sub { |
|
|
1319 | my ($ports, $add, $chg, $del) = @_; |
|
|
1320 | $ports = [keys %$ports]; |
|
|
1321 | |
|
|
1322 | # now $ports, $add and $del are the same as |
|
|
1323 | # were originally passed by grp_mon. |
|
|
1324 | ... |
|
|
1325 | }; |
|
|
1326 | |
|
|
1327 | =item Nodes not longer connect to all other nodes. |
|
|
1328 | |
|
|
1329 | In AEMP 1.x, every node automatically loads the L<AnyEvent::MP::Global> |
|
|
1330 | module, which in turn would create connections to all other nodes in the |
|
|
1331 | network (helped by the seed nodes). |
|
|
1332 | |
|
|
1333 | In version 2.x, global nodes still connect to all other global nodes, but |
|
|
1334 | other nodes don't - now every node either is a global node itself, or |
|
|
1335 | attaches itself to another global node. |
|
|
1336 | |
|
|
1337 | If a node isn't a global node itself, then it attaches itself to one |
|
|
1338 | of its seed nodes. If that seed node isn't a global node yet, it will |
|
|
1339 | automatically be upgraded to a global node. |
|
|
1340 | |
|
|
1341 | So in many cases, nothing needs to be changed - one just has to make sure |
|
|
1342 | that all seed nodes are meshed together with the other seed nodes (as with |
|
|
1343 | AEMP 1.x), and other nodes specify them as seed nodes. This is most easily |
|
|
1344 | achieved by specifying the same set of seed nodes for all nodes in the |
|
|
1345 | network. |
|
|
1346 | |
|
|
1347 | Not opening a connection to every other node is usually an advantage, |
|
|
1348 | except when you need the lower latency of an already established |
|
|
1349 | connection. To ensure a node establishes a connection to another node, |
|
|
1350 | you can monitor the node port (C<mon $node, ...>), which will attempt to |
|
|
1351 | create the connection (and notify you when the connection fails). |
|
|
1352 | |
|
|
1353 | =item Listener-less nodes (nodes without binds) are gone. |
|
|
1354 | |
|
|
1355 | And are not coming back, at least not in their old form. If no C<binds> |
|
|
1356 | are specified for a node, AnyEvent::MP assumes a default of C<*:*>. |
|
|
1357 | |
|
|
1358 | There are vague plans to implement some form of routing domains, which |
|
|
1359 | might or might not bring back listener-less nodes, but don't count on it. |
|
|
1360 | |
|
|
1361 | The fact that most connections are now optional somewhat mitigates this, |
|
|
1362 | as a node can be effectively unreachable from the outside without any |
|
|
1363 | problems, as long as it isn't a global node and only reaches out to other |
|
|
1364 | nodes (as opposed to being contacted from other nodes). |
|
|
1365 | |
|
|
1366 | =item $AnyEvent::MP::Kernel::WARN has gone. |
|
|
1367 | |
|
|
1368 | AnyEvent has acquired a logging framework (L<AnyEvent::Log>), and AEMP now |
|
|
1369 | uses this, and so should your programs. |
|
|
1370 | |
|
|
1371 | Every module now documents what kinds of messages it generates, with |
|
|
1372 | AnyEvent::MP acting as a catch all. |
|
|
1373 | |
|
|
1374 | On the positive side, this means that instead of setting |
|
|
1375 | C<PERL_ANYEVENT_MP_WARNLEVEL>, you can get away by setting C<AE_VERBOSE> - |
|
|
1376 | much less to type. |
|
|
1377 | |
|
|
1378 | =back |
|
|
1379 | |
|
|
1380 | =head1 LOGGING |
|
|
1381 | |
|
|
1382 | AnyEvent::MP does not normally log anything by itself, but since it is the |
|
|
1383 | root of the context hierarchy for AnyEvent::MP modules, it will receive |
|
|
1384 | all log messages by submodules. |
|
|
1385 | |
872 | =head1 SEE ALSO |
1386 | =head1 SEE ALSO |
|
|
1387 | |
|
|
1388 | L<AnyEvent::MP::Intro> - a gentle introduction. |
|
|
1389 | |
|
|
1390 | L<AnyEvent::MP::Kernel> - more, lower-level, stuff. |
|
|
1391 | |
|
|
1392 | L<AnyEvent::MP::Global> - network maintenance and port groups, to find |
|
|
1393 | your applications. |
|
|
1394 | |
|
|
1395 | L<AnyEvent::MP::DataConn> - establish data connections between nodes. |
|
|
1396 | |
|
|
1397 | L<AnyEvent::MP::LogCatcher> - simple service to display log messages from |
|
|
1398 | all nodes. |
873 | |
1399 | |
874 | L<AnyEvent>. |
1400 | L<AnyEvent>. |
875 | |
1401 | |
876 | =head1 AUTHOR |
1402 | =head1 AUTHOR |
877 | |
1403 | |