… | |
… | |
35 | # destroy a port again |
35 | # destroy a port again |
36 | kil $port; # "normal" kill |
36 | kil $port; # "normal" kill |
37 | kil $port, my_error => "everything is broken"; # error kill |
37 | kil $port, my_error => "everything is broken"; # error kill |
38 | |
38 | |
39 | # monitoring |
39 | # monitoring |
40 | mon $localport, $cb->(@msg) # callback is invoked on death |
40 | mon $port, $cb->(@msg) # callback is invoked on death |
41 | mon $localport, $otherport # kill otherport on abnormal death |
41 | mon $port, $localport # kill localport on abnormal death |
42 | mon $localport, $otherport, @msg # send message on death |
42 | mon $port, $localport, @msg # send message on death |
43 | |
43 | |
44 | # temporarily execute code in port context |
44 | # temporarily execute code in port context |
45 | peval $port, sub { die "kill the port!" }; |
45 | peval $port, sub { die "kill the port!" }; |
46 | |
46 | |
47 | # execute callbacks in $SELF port context |
47 | # execute callbacks in $SELF port context |
… | |
… | |
82 | |
82 | |
83 | Ports are represented by (printable) strings called "port IDs". |
83 | Ports are represented by (printable) strings called "port IDs". |
84 | |
84 | |
85 | =item port ID - C<nodeid#portname> |
85 | =item port ID - C<nodeid#portname> |
86 | |
86 | |
87 | A port ID is the concatenation of a node ID, a hash-mark (C<#>) as |
87 | A port ID is the concatenation of a node ID, a hash-mark (C<#>) |
88 | separator, and a port name (a printable string of unspecified format). |
88 | as separator, and a port name (a printable string of unspecified |
|
|
89 | format created by AnyEvent::MP). |
89 | |
90 | |
90 | =item node |
91 | =item node |
91 | |
92 | |
92 | A node is a single process containing at least one port - the node port, |
93 | A node is a single process containing at least one port - the node port, |
93 | which enables nodes to manage each other remotely, and to create new |
94 | which enables nodes to manage each other remotely, and to create new |
… | |
… | |
175 | |
176 | |
176 | =cut |
177 | =cut |
177 | |
178 | |
178 | package AnyEvent::MP; |
179 | package AnyEvent::MP; |
179 | |
180 | |
|
|
181 | use AnyEvent::MP::Config (); |
180 | use AnyEvent::MP::Kernel; |
182 | use AnyEvent::MP::Kernel; |
|
|
183 | use AnyEvent::MP::Kernel qw(%NODE %PORT %PORT_DATA $UNIQ $RUNIQ $ID); |
181 | |
184 | |
182 | use common::sense; |
185 | use common::sense; |
183 | |
186 | |
184 | use Carp (); |
187 | use Carp (); |
185 | |
188 | |
186 | use AE (); |
189 | use AE (); |
|
|
190 | use Guard (); |
187 | |
191 | |
188 | use base "Exporter"; |
192 | use base "Exporter"; |
189 | |
193 | |
190 | our $VERSION = '1.30'; |
194 | our $VERSION = $AnyEvent::MP::Config::VERSION; |
191 | |
195 | |
192 | our @EXPORT = qw( |
196 | our @EXPORT = qw( |
193 | NODE $NODE *SELF node_of after |
197 | NODE $NODE *SELF node_of after |
194 | configure |
198 | configure |
195 | snd rcv mon mon_guard kil psub peval spawn cal |
199 | snd rcv mon mon_guard kil psub peval spawn cal |
196 | port |
200 | port |
|
|
201 | db_set db_del db_reg |
|
|
202 | db_mon db_family db_keys db_values |
197 | ); |
203 | ); |
198 | |
204 | |
199 | our $SELF; |
205 | our $SELF; |
200 | |
206 | |
201 | sub _self_die() { |
207 | sub _self_die() { |
… | |
… | |
221 | Before a node can talk to other nodes on the network (i.e. enter |
227 | Before a node can talk to other nodes on the network (i.e. enter |
222 | "distributed mode") it has to configure itself - the minimum a node needs |
228 | "distributed mode") it has to configure itself - the minimum a node needs |
223 | to know is its own name, and optionally it should know the addresses of |
229 | to know is its own name, and optionally it should know the addresses of |
224 | some other nodes in the network to discover other nodes. |
230 | some other nodes in the network to discover other nodes. |
225 | |
231 | |
226 | The key/value pairs are basically the same ones as documented for the |
|
|
227 | F<aemp> command line utility (sans the set/del prefix). |
|
|
228 | |
|
|
229 | This function configures a node - it must be called exactly once (or |
232 | This function configures a node - it must be called exactly once (or |
230 | never) before calling other AnyEvent::MP functions. |
233 | never) before calling other AnyEvent::MP functions. |
|
|
234 | |
|
|
235 | The key/value pairs are basically the same ones as documented for the |
|
|
236 | F<aemp> command line utility (sans the set/del prefix), with these additions: |
|
|
237 | |
|
|
238 | =over 4 |
|
|
239 | |
|
|
240 | =item norc => $boolean (default false) |
|
|
241 | |
|
|
242 | If true, then the rc file (e.g. F<~/.perl-anyevent-mp>) will I<not> |
|
|
243 | be consulted - all configuraiton options must be specified in the |
|
|
244 | C<configure> call. |
|
|
245 | |
|
|
246 | =item force => $boolean (default false) |
|
|
247 | |
|
|
248 | IF true, then the values specified in the C<configure> will take |
|
|
249 | precedence over any values configured via the rc file. The default is for |
|
|
250 | the rc file to override any options specified in the program. |
|
|
251 | |
|
|
252 | =item secure => $pass->($nodeid) |
|
|
253 | |
|
|
254 | In addition to specifying a boolean, you can specify a code reference that |
|
|
255 | is called for every remote execution attempt - the execution request is |
|
|
256 | granted iff the callback returns a true value. |
|
|
257 | |
|
|
258 | See F<semp setsecure> for more info. |
|
|
259 | |
|
|
260 | =back |
231 | |
261 | |
232 | =over 4 |
262 | =over 4 |
233 | |
263 | |
234 | =item step 1, gathering configuration from profiles |
264 | =item step 1, gathering configuration from profiles |
235 | |
265 | |
… | |
… | |
249 | That means that the values specified in the profile have highest priority |
279 | That means that the values specified in the profile have highest priority |
250 | and the values specified directly via C<configure> have lowest priority, |
280 | and the values specified directly via C<configure> have lowest priority, |
251 | and can only be used to specify defaults. |
281 | and can only be used to specify defaults. |
252 | |
282 | |
253 | If the profile specifies a node ID, then this will become the node ID of |
283 | If the profile specifies a node ID, then this will become the node ID of |
254 | this process. If not, then the profile name will be used as node ID. The |
284 | this process. If not, then the profile name will be used as node ID, with |
255 | special node ID of C<anon/> will be replaced by a random node ID. |
285 | a unique randoms tring (C</%u>) appended. |
|
|
286 | |
|
|
287 | The node ID can contain some C<%> sequences that are expanded: C<%n> |
|
|
288 | is expanded to the local nodename, C<%u> is replaced by a random |
|
|
289 | strign to make the node unique. For example, the F<aemp> commandline |
|
|
290 | utility uses C<aemp/%n/%u> as nodename, which might expand to |
|
|
291 | C<aemp/cerebro/ZQDGSIkRhEZQDGSIkRhE>. |
256 | |
292 | |
257 | =item step 2, bind listener sockets |
293 | =item step 2, bind listener sockets |
258 | |
294 | |
259 | The next step is to look up the binds in the profile, followed by binding |
295 | The next step is to look up the binds in the profile, followed by binding |
260 | aemp protocol listeners on all binds specified (it is possible and valid |
296 | aemp protocol listeners on all binds specified (it is possible and valid |
… | |
… | |
277 | Example: become a distributed node using the local node name as profile. |
313 | Example: become a distributed node using the local node name as profile. |
278 | This should be the most common form of invocation for "daemon"-type nodes. |
314 | This should be the most common form of invocation for "daemon"-type nodes. |
279 | |
315 | |
280 | configure |
316 | configure |
281 | |
317 | |
282 | Example: become an anonymous node. This form is often used for commandline |
318 | Example: become a semi-anonymous node. This form is often used for |
283 | clients. |
319 | commandline clients. |
284 | |
320 | |
285 | configure nodeid => "anon/"; |
321 | configure nodeid => "myscript/%n/%u"; |
286 | |
322 | |
287 | Example: configure a node using a profile called seed, which si suitable |
323 | Example: configure a node using a profile called seed, which is suitable |
288 | for a seed node as it binds on all local addresses on a fixed port (4040, |
324 | for a seed node as it binds on all local addresses on a fixed port (4040, |
289 | customary for aemp). |
325 | customary for aemp). |
290 | |
326 | |
291 | # use the aemp commandline utility |
327 | # use the aemp commandline utility |
292 | # aemp profile seed nodeid anon/ binds '*:4040' |
328 | # aemp profile seed binds '*:4040' |
293 | |
329 | |
294 | # then use it |
330 | # then use it |
295 | configure profile => "seed"; |
331 | configure profile => "seed"; |
296 | |
332 | |
297 | # or simply use aemp from the shell again: |
333 | # or simply use aemp from the shell again: |
… | |
… | |
362 | |
398 | |
363 | =cut |
399 | =cut |
364 | |
400 | |
365 | sub rcv($@); |
401 | sub rcv($@); |
366 | |
402 | |
367 | sub _kilme { |
403 | my $KILME = sub { |
368 | die "received message on port without callback"; |
404 | die "received message on port without callback"; |
369 | } |
405 | }; |
370 | |
406 | |
371 | sub port(;&) { |
407 | sub port(;&) { |
372 | my $id = "$UNIQ." . $ID++; |
408 | my $id = $UNIQ . ++$ID; |
373 | my $port = "$NODE#$id"; |
409 | my $port = "$NODE#$id"; |
374 | |
410 | |
375 | rcv $port, shift || \&_kilme; |
411 | rcv $port, shift || $KILME; |
376 | |
412 | |
377 | $port |
413 | $port |
378 | } |
414 | } |
379 | |
415 | |
380 | =item rcv $local_port, $callback->(@msg) |
416 | =item rcv $local_port, $callback->(@msg) |
… | |
… | |
650 | } |
686 | } |
651 | |
687 | |
652 | $node->monitor ($port, $cb); |
688 | $node->monitor ($port, $cb); |
653 | |
689 | |
654 | defined wantarray |
690 | defined wantarray |
655 | and ($cb += 0, AnyEvent::Util::guard { $node->unmonitor ($port, $cb) }) |
691 | and ($cb += 0, Guard::guard { $node->unmonitor ($port, $cb) }) |
656 | } |
692 | } |
657 | |
693 | |
658 | =item $guard = mon_guard $port, $ref, $ref... |
694 | =item $guard = mon_guard $port, $ref, $ref... |
659 | |
695 | |
660 | Monitors the given C<$port> and keeps the passed references. When the port |
696 | Monitors the given C<$port> and keeps the passed references. When the port |
… | |
… | |
764 | } |
800 | } |
765 | |
801 | |
766 | sub spawn(@) { |
802 | sub spawn(@) { |
767 | my ($nodeid, undef) = split /#/, shift, 2; |
803 | my ($nodeid, undef) = split /#/, shift, 2; |
768 | |
804 | |
769 | my $id = "$RUNIQ." . $ID++; |
805 | my $id = $RUNIQ . ++$ID; |
770 | |
806 | |
771 | $_[0] =~ /::/ |
807 | $_[0] =~ /::/ |
772 | or Carp::croak "spawn init function must be a fully-qualified name, caught"; |
808 | or Carp::croak "spawn init function must be a fully-qualified name, caught"; |
773 | |
809 | |
774 | snd_to_func $nodeid, "AnyEvent::MP::_spawn" => $id, @_; |
810 | snd_to_func $nodeid, "AnyEvent::MP::_spawn" => $id, @_; |
775 | |
811 | |
776 | "$nodeid#$id" |
812 | "$nodeid#$id" |
777 | } |
813 | } |
|
|
814 | |
778 | |
815 | |
779 | =item after $timeout, @msg |
816 | =item after $timeout, @msg |
780 | |
817 | |
781 | =item after $timeout, $callback |
818 | =item after $timeout, $callback |
782 | |
819 | |
… | |
… | |
797 | ref $action[0] |
834 | ref $action[0] |
798 | ? $action[0]() |
835 | ? $action[0]() |
799 | : snd @action; |
836 | : snd @action; |
800 | }; |
837 | }; |
801 | } |
838 | } |
|
|
839 | |
|
|
840 | #=item $cb2 = timeout $seconds, $cb[, @args] |
802 | |
841 | |
803 | =item cal $port, @msg, $callback[, $timeout] |
842 | =item cal $port, @msg, $callback[, $timeout] |
804 | |
843 | |
805 | A simple form of RPC - sends a message to the given C<$port> with the |
844 | A simple form of RPC - sends a message to the given C<$port> with the |
806 | given contents (C<@msg>), but adds a reply port to the message. |
845 | given contents (C<@msg>), but adds a reply port to the message. |
… | |
… | |
852 | $port |
891 | $port |
853 | } |
892 | } |
854 | |
893 | |
855 | =back |
894 | =back |
856 | |
895 | |
|
|
896 | =head1 DISTRIBUTED DATABASE |
|
|
897 | |
|
|
898 | AnyEvent::MP comes with a simple distributed database. The database will |
|
|
899 | be mirrored asynchronously on all global nodes. Other nodes bind to one |
|
|
900 | of the global nodes for their needs. Every node has a "local database" |
|
|
901 | which contains all the values that are set locally. All local databases |
|
|
902 | are merged together to form the global database, which can be queried. |
|
|
903 | |
|
|
904 | The database structure is that of a two-level hash - the database hash |
|
|
905 | contains hashes which contain values, similarly to a perl hash of hashes, |
|
|
906 | i.e.: |
|
|
907 | |
|
|
908 | $DATABASE{$family}{$subkey} = $value |
|
|
909 | |
|
|
910 | The top level hash key is called "family", and the second-level hash key |
|
|
911 | is called "subkey" or simply "key". |
|
|
912 | |
|
|
913 | The family must be alphanumeric, i.e. start with a letter and consist |
|
|
914 | of letters, digits, underscores and colons (C<[A-Za-z][A-Za-z0-9_:]*>, |
|
|
915 | pretty much like Perl module names. |
|
|
916 | |
|
|
917 | As the family namespace is global, it is recommended to prefix family names |
|
|
918 | with the name of the application or module using it. |
|
|
919 | |
|
|
920 | The subkeys must be non-empty strings, with no further restrictions. |
|
|
921 | |
|
|
922 | The values should preferably be strings, but other perl scalars should |
|
|
923 | work as well (such as C<undef>, arrays and hashes). |
|
|
924 | |
|
|
925 | Every database entry is owned by one node - adding the same family/subkey |
|
|
926 | combination on multiple nodes will not cause discomfort for AnyEvent::MP, |
|
|
927 | but the result might be nondeterministic, i.e. the key might have |
|
|
928 | different values on different nodes. |
|
|
929 | |
|
|
930 | Different subkeys in the same family can be owned by different nodes |
|
|
931 | without problems, and in fact, this is the common method to create worker |
|
|
932 | pools. For example, a worker port for image scaling might do this: |
|
|
933 | |
|
|
934 | db_set my_image_scalers => $port; |
|
|
935 | |
|
|
936 | And clients looking for an image scaler will want to get the |
|
|
937 | C<my_image_scalers> keys from time to time: |
|
|
938 | |
|
|
939 | db_keys my_image_scalers => sub { |
|
|
940 | @ports = @{ $_[0] }; |
|
|
941 | }; |
|
|
942 | |
|
|
943 | Or better yet, they want to monitor the database family, so they always |
|
|
944 | have a reasonable up-to-date copy: |
|
|
945 | |
|
|
946 | db_mon my_image_scalers => sub { |
|
|
947 | @ports = keys %{ $_[0] }; |
|
|
948 | }; |
|
|
949 | |
|
|
950 | In general, you can set or delete single subkeys, but query and monitor |
|
|
951 | whole families only. |
|
|
952 | |
|
|
953 | If you feel the need to monitor or query a single subkey, try giving it |
|
|
954 | it's own family. |
|
|
955 | |
|
|
956 | =over |
|
|
957 | |
|
|
958 | =item db_set $family => $subkey [=> $value] |
|
|
959 | |
|
|
960 | Sets (or replaces) a key to the database - if C<$value> is omitted, |
|
|
961 | C<undef> is used instead. |
|
|
962 | |
|
|
963 | =item db_del $family => $subkey... |
|
|
964 | |
|
|
965 | Deletes one or more subkeys from the database family. |
|
|
966 | |
|
|
967 | =item $guard = db_reg $family => $subkey [=> $value] |
|
|
968 | |
|
|
969 | Sets the key on the database and returns a guard. When the guard is |
|
|
970 | destroyed, the key is deleted from the database. If C<$value> is missing, |
|
|
971 | then C<undef> is used. |
|
|
972 | |
|
|
973 | =item db_family $family => $cb->(\%familyhash) |
|
|
974 | |
|
|
975 | Queries the named database C<$family> and call the callback with the |
|
|
976 | family represented as a hash. You can keep and freely modify the hash. |
|
|
977 | |
|
|
978 | =item db_keys $family => $cb->(\@keys) |
|
|
979 | |
|
|
980 | Same as C<db_family>, except it only queries the family I<subkeys> and passes |
|
|
981 | them as array reference to the callback. |
|
|
982 | |
|
|
983 | =item db_values $family => $cb->(\@values) |
|
|
984 | |
|
|
985 | Same as C<db_family>, except it only queries the family I<values> and passes them |
|
|
986 | as array reference to the callback. |
|
|
987 | |
|
|
988 | =item $guard = db_mon $family => $cb->($familyhash, \@added, \@changed, \@deleted) |
|
|
989 | |
|
|
990 | Creates a monitor on the given database family. Each time a key is set |
|
|
991 | or or is deleted the callback is called with a hash containing the |
|
|
992 | database family and three lists of added, changed and deleted subkeys, |
|
|
993 | respectively. If no keys have changed then the array reference might be |
|
|
994 | C<undef> or even missing. |
|
|
995 | |
|
|
996 | If not called in void context, a guard object is returned that, when |
|
|
997 | destroyed, stops the monitor. |
|
|
998 | |
|
|
999 | The family hash reference and the key arrays belong to AnyEvent::MP and |
|
|
1000 | B<must not be modified or stored> by the callback. When in doubt, make a |
|
|
1001 | copy. |
|
|
1002 | |
|
|
1003 | As soon as possible after the monitoring starts, the callback will be |
|
|
1004 | called with the intiial contents of the family, even if it is empty, |
|
|
1005 | i.e. there will always be a timely call to the callback with the current |
|
|
1006 | contents. |
|
|
1007 | |
|
|
1008 | It is possible that the callback is called with a change event even though |
|
|
1009 | the subkey is already present and the value has not changed. |
|
|
1010 | |
|
|
1011 | The monitoring stops when the guard object is destroyed. |
|
|
1012 | |
|
|
1013 | Example: on every change to the family "mygroup", print out all keys. |
|
|
1014 | |
|
|
1015 | my $guard = db_mon mygroup => sub { |
|
|
1016 | my ($family, $a, $c, $d) = @_; |
|
|
1017 | print "mygroup members: ", (join " ", keys %$family), "\n"; |
|
|
1018 | }; |
|
|
1019 | |
|
|
1020 | Exmaple: wait until the family "My::Module::workers" is non-empty. |
|
|
1021 | |
|
|
1022 | my $guard; $guard = db_mon My::Module::workers => sub { |
|
|
1023 | my ($family, $a, $c, $d) = @_; |
|
|
1024 | return unless %$family; |
|
|
1025 | undef $guard; |
|
|
1026 | print "My::Module::workers now nonempty\n"; |
|
|
1027 | }; |
|
|
1028 | |
|
|
1029 | Example: print all changes to the family "AnyRvent::Fantasy::Module". |
|
|
1030 | |
|
|
1031 | my $guard = db_mon AnyRvent::Fantasy::Module => sub { |
|
|
1032 | my ($family, $a, $c, $d) = @_; |
|
|
1033 | |
|
|
1034 | print "+$_=$family->{$_}\n" for @$a; |
|
|
1035 | print "*$_=$family->{$_}\n" for @$c; |
|
|
1036 | print "-$_=$family->{$_}\n" for @$d; |
|
|
1037 | }; |
|
|
1038 | |
|
|
1039 | =cut |
|
|
1040 | |
|
|
1041 | =back |
|
|
1042 | |
857 | =head1 AnyEvent::MP vs. Distributed Erlang |
1043 | =head1 AnyEvent::MP vs. Distributed Erlang |
858 | |
1044 | |
859 | AnyEvent::MP got lots of its ideas from distributed Erlang (Erlang node |
1045 | AnyEvent::MP got lots of its ideas from distributed Erlang (Erlang node |
860 | == aemp node, Erlang process == aemp port), so many of the documents and |
1046 | == aemp node, Erlang process == aemp port), so many of the documents and |
861 | programming techniques employed by Erlang apply to AnyEvent::MP. Here is a |
1047 | programming techniques employed by Erlang apply to AnyEvent::MP. Here is a |