1 package DBIx::Class::Storage::DBI::Replicated;
7 require DBIx::Class::Optional::Dependencies;
8 if ( my $missing = DBIx::Class::Optional::Dependencies->req_missing_for('replicated') ) {
9 die "The following modules are required for Replicated storage support: $missing\n";
14 use DBIx::Class::Storage::DBI;
15 use DBIx::Class::Storage::DBI::Replicated::Pool;
16 use DBIx::Class::Storage::DBI::Replicated::Balancer;
17 use DBIx::Class::Storage::DBI::Replicated::Types qw/BalancerClassNamePart DBICSchema DBICStorageDBI/;
18 use MooseX::Types::Moose qw/ClassName HashRef Object/;
19 use Scalar::Util 'reftype';
21 use List::Util qw( min max );
22 use Context::Preserve 'preserve_context';
24 use namespace::clean -except => 'meta';
28 DBIx::Class::Storage::DBI::Replicated - BETA Replicated database support
32 The Following example shows how to change an existing $schema to a replicated
33 storage type, add some replicated (read-only) databases, and perform reporting
36 You should set the 'storage_type attribute to a replicated type. You should
37 also define your arguments, such as which balancer you want and any arguments
38 that the Pool object should get.
40 my $schema = Schema::Class->clone;
41 $schema->storage_type(['::DBI::Replicated', { balancer_type => '::Random' }]);
42 $schema->connection(...);
44 Next, you need to add in the Replicants. Basically this is an array of
45 arrayrefs, where each arrayref is database connect information. Think of these
46 arguments as what you'd pass to the 'normal' $schema->connect method.
48 $schema->storage->connect_replicants(
49 [$dsn1, $user, $pass, \%opts],
50 [$dsn2, $user, $pass, \%opts],
51 [$dsn3, $user, $pass, \%opts],
54 Now, just use the $schema as you normally would. Automatically all reads will
55 be delegated to the replicants, while writes to the master.
57 $schema->resultset('Source')->search({name=>'etc'});
59 You can force a given query to use a particular storage using the search
60 attribute 'force_pool'. For example:
62 my $rs = $schema->resultset('Source')->search(undef, {force_pool=>'master'});
64 Now $rs will force everything (both reads and writes) to use whatever was setup
65 as the master storage. 'master' is hardcoded to always point to the Master,
66 but you can also use any Replicant name. Please see:
67 L<DBIx::Class::Storage::DBI::Replicated::Pool> and the replicants attribute for more.
69 Also see transactions and L</execute_reliably> for alternative ways to
70 force read traffic to the master. In general, you should wrap your statements
71 in a transaction when you are reading and writing to the same tables at the
72 same time, since your replicants will often lag a bit behind the master.
74 If you have a multi-statement read only transaction you can force it to select
75 a random server in the pool by:
77 my $rs = $schema->resultset('Source')->search( undef,
78 { force_pool => $db->storage->read_handler->next_storage }
83 Warning: This class is marked BETA. This has been running a production
84 website using MySQL native replication as its backend and we have some decent
85 test coverage but the code hasn't yet been stressed by a variety of databases.
86 Individual DBs may have quirks we are not aware of. Please use this in first
87 development and pass along your experiences/bug fixes.
89 This class implements replicated data store for DBI. Currently you can define
90 one master and numerous slave database connections. All write-type queries
91 (INSERT, UPDATE, DELETE and even LAST_INSERT_ID) are routed to master
92 database, all read-type queries (SELECTs) go to the slave database.
94 Basically, any method request that L<DBIx::Class::Storage::DBI> would normally
95 handle gets delegated to one of the two attributes: L</read_handler> or to
96 L</write_handler>. Additionally, some methods need to be distributed
97 to all existing storages. This way our storage class is a drop in replacement
98 for L<DBIx::Class::Storage::DBI>.
100 Read traffic is spread across the replicants (slaves) occurring to a user
101 selected algorithm. The default algorithm is random weighted.
105 The consistency between master and replicants is database specific. The Pool
106 gives you a method to validate its replicants, removing and replacing them
107 when they fail/pass predefined criteria. Please make careful use of the ways
108 to force a query to run against Master when needed.
112 Replicated Storage has additional requirements not currently part of
113 L<DBIx::Class>. See L<DBIx::Class::Optional::Dependencies> for more details.
117 This class defines the following attributes.
121 The underlying L<DBIx::Class::Schema> object this storage is attaching
134 Contains the classname which will instantiate the L</pool> object. Defaults
135 to: L<DBIx::Class::Storage::DBI::Replicated::Pool>.
142 default=>'DBIx::Class::Storage::DBI::Replicated::Pool',
144 'create_pool' => 'new',
150 Contains a hashref of initialized information to pass to the Balancer object.
151 See L<DBIx::Class::Storage::DBI::Replicated::Pool> for available arguments.
165 The replication pool requires a balance class to provider the methods for
166 choose how to spread the query load across each replicant in the pool.
170 has 'balancer_type' => (
172 isa=>BalancerClassNamePart,
175 default=> 'DBIx::Class::Storage::DBI::Replicated::Balancer::First',
177 'create_balancer' => 'new',
183 Contains a hashref of initialized information to pass to the Balancer object.
184 See L<DBIx::Class::Storage::DBI::Replicated::Balancer> for available arguments.
188 has 'balancer_args' => (
198 Is a L<DBIx::Class::Storage::DBI::Replicated::Pool> or derived class. This is a
199 container class for one or more replicated databases.
205 isa=>'DBIx::Class::Storage::DBI::Replicated::Pool',
216 Is a L<DBIx::Class::Storage::DBI::Replicated::Balancer> or derived class. This
217 is a class that takes a pool (L<DBIx::Class::Storage::DBI::Replicated::Pool>)
223 isa=>'DBIx::Class::Storage::DBI::Replicated::Balancer',
225 handles=>[qw/auto_validate_every/],
230 The master defines the canonical state for a pool of connected databases. All
231 the replicants are expected to match this databases state. Thus, in a classic
232 Master / Slaves distributed system, all the slaves are expected to replicate
233 the Master's state as quick as possible. This is the only database in the
234 pool of databases that is allowed to handle write traffic.
244 =head1 ATTRIBUTES IMPLEMENTING THE DBIx::Storage::DBI INTERFACE
246 The following methods are delegated all the methods required for the
247 L<DBIx::Class::Storage::DBI> interface.
251 my $method_dispatch = {
263 deployment_statements
266 build_datetime_parser
281 with_deferred_fk_checks
289 relname_to_table_alias
291 _default_dbi_connect_attributes
293 _dbic_connect_attributes
299 bind_attribute_by_data_type
303 _dbh_execute_for_fetch
305 _dbh_execute_inserts_with_no_binds
306 _select_args_to_query
309 _normalize_connect_info
324 /, Class::MOP::Class->initialize('DBIx::Class::Storage::DBIHacks')->get_method_list ],
329 _dbh_columns_info_for
332 unimplemented => [qw/
333 _arm_global_destructor
337 get_use_dbms_capability
338 set_use_dbms_capability
347 _determine_connector_driver
348 _extract_driver_from_connect_info
350 _warn_undetermined_driver
357 _perform_autoinc_retrieval
358 _autoinc_supplied_for_op
371 # the capability framework
372 # not sure if CMOP->initialize does evil things to DBIC::S::DBI, fix if a problem
374 { $_ =~ /^ _ (?: use | supports | determine_supports ) _ /x and $_ ne '_use_multicolumn_in' }
375 ( Class::MOP::Class->initialize('DBIx::Class::Storage::DBI')->get_all_method_names )
379 # this only happens during DBIC-internal testing
380 if ( $INC{"t/lib/ANFANG.pm"} ) {
383 for my $type (keys %$method_dispatch) {
384 for (@{$method_dispatch->{$type}}) {
385 push @{$seen->{$_}}, $type;
389 if (my @dupes = grep { @{$seen->{$_}} > 1 } keys %$seen) {
391 'The following methods show up multiple times in ::Storage::DBI::Replicated handlers:',
392 (map { "$_: " . (join ', ', @{$seen->{$_}}) } sort @dupes),
397 if (my @cant = grep { ! DBIx::Class::Storage::DBI->can($_) } keys %$seen) {
399 '::Storage::DBI::Replicated specifies handling of the following *NON EXISTING* ::Storage::DBI methods:',
406 for my $method (@{$method_dispatch->{unimplemented}}) {
407 __PACKAGE__->meta->add_method($method, sub {
409 $self->throw_exception(
410 "$method() may not be called on '@{[ blessed $self ]}' objects, "
411 . 'call it on a specific pool instance instead'
418 Defines an object that implements the read side of L<DBIx::Class::Storage::DBI>.
422 has 'read_handler' => (
426 handles=>$method_dispatch->{reader},
431 Defines an object that implements the write side of L<DBIx::Class::Storage::DBI>,
432 as well as methods that don't write or read that can be called on only one
433 storage, methods that return a C<$dbh>, and any methods that don't make sense to
438 has 'write_handler' => (
442 handles=>$method_dispatch->{writer},
447 has _master_connect_info_opts =>
448 (is => 'rw', isa => HashRef, default => sub { {} });
450 =head2 around: connect_info
452 Preserves master's C<connect_info> options (for merging with replicants.)
453 Also sets any Replicated-related options from connect_info, such as
454 C<pool_type>, C<pool_args>, C<balancer_type> and C<balancer_args>.
458 around connect_info => sub {
459 my ($next, $self, $info, @extra) = @_;
461 $self->throw_exception(
462 'connect_info can not be retrieved from a replicated storage - '
463 . 'accessor must be called on a specific pool instance'
464 ) unless defined $info;
466 my $merge = Hash::Merge->new('LEFT_PRECEDENT');
469 for my $arg (@$info) {
470 next unless (reftype($arg)||'') eq 'HASH';
471 %opts = %{ $merge->merge($arg, \%opts) };
475 if (@opts{qw/pool_type pool_args/}) {
476 $self->pool_type(delete $opts{pool_type})
480 $merge->merge((delete $opts{pool_args} || {}), $self->pool_args)
483 ## Since we possibly changed the pool_args, we need to clear the current
484 ## pool object so that next time it is used it will be rebuilt.
488 if (@opts{qw/balancer_type balancer_args/}) {
489 $self->balancer_type(delete $opts{balancer_type})
490 if $opts{balancer_type};
492 $self->balancer_args(
493 $merge->merge((delete $opts{balancer_args} || {}), $self->balancer_args)
496 $self->balancer($self->_build_balancer)
500 $self->_master_connect_info_opts(\%opts);
502 return preserve_context {
503 $self->$next($info, @extra);
505 # Make sure master is blessed into the correct class and apply role to it.
506 my $master = $self->master;
507 $master->_determine_driver;
508 Moose::Meta::Class->initialize(ref $master);
510 DBIx::Class::Storage::DBI::Replicated::WithDSN->meta->apply($master);
512 # link pool back to master
513 $self->pool->master($master);
519 This class defines the following methods.
523 L<DBIx::Class::Schema> when instantiating its storage passed itself as the
524 first argument. So we need to massage the arguments a bit so that all the
525 bits get put into the correct places.
530 my ($class, $schema, $storage_type_args, @args) = @_;
541 Lazy builder for the L</master> attribute.
547 my $master = DBIx::Class::Storage::DBI->new($self->schema);
553 Lazy builder for the L</pool> attribute.
559 $self->create_pool(%{$self->pool_args});
562 =head2 _build_balancer
564 Lazy builder for the L</balancer> attribute. This takes a Pool object so that
565 the balancer knows which pool it's balancing.
569 sub _build_balancer {
571 $self->create_balancer(
573 master=>$self->master,
574 %{$self->balancer_args},
578 =head2 _build_write_handler
580 Lazy builder for the L</write_handler> attribute. The default is to set this to
585 sub _build_write_handler {
586 return shift->master;
589 =head2 _build_read_handler
591 Lazy builder for the L</read_handler> attribute. The default is to set this to
596 sub _build_read_handler {
597 return shift->balancer;
600 =head2 around: connect_replicants
602 All calls to connect_replicants needs to have an existing $schema tacked onto
603 top of the args, since L<DBIx::Class::Storage::DBI> needs it, and any
604 L<connect_info|DBIx::Class::Storage::DBI/connect_info>
605 options merged with the master, with replicant opts having higher priority.
609 around connect_replicants => sub {
610 my ($next, $self, @args) = @_;
613 $r = [ $r ] unless reftype $r eq 'ARRAY';
615 $self->throw_exception('coderef replicant connect_info not supported')
616 if ref $r->[0] && reftype $r->[0] eq 'CODE';
618 # any connect_info options?
620 $i++ while $i < @$r && (reftype($r->[$i])||'') ne 'HASH';
623 $r->[$i] = {} unless $r->[$i];
625 # merge if two hashes
626 my @hashes = @$r[$i .. $#{$r}];
628 $self->throw_exception('invalid connect_info options')
629 if (grep { reftype($_) eq 'HASH' } @hashes) != @hashes;
631 $self->throw_exception('too many hashrefs in connect_info')
634 my $merge = Hash::Merge->new('LEFT_PRECEDENT');
635 my %opts = %{ $merge->merge(reverse @hashes) };
638 splice @$r, $i+1, ($#{$r} - $i), ();
640 # make sure master/replicants opts don't clash
641 my %master_opts = %{ $self->_master_connect_info_opts };
642 if (exists $opts{dbh_maker}) {
643 delete @master_opts{qw/dsn user password/};
645 delete $master_opts{dbh_maker};
648 %opts = %{ $merge->merge(\%opts, \%master_opts) };
654 $self->$next($self->schema, @args);
659 Returns an array of all the connected storage backends. The first element
660 in the returned array is the master, and the rest are each of the
667 return grep {defined $_ && blessed $_} (
669 values %{ $self->replicants },
673 =head2 execute_reliably ($coderef, ?@args)
675 Given a coderef, saves the current state of the L</read_handler>, forces it to
676 use reliable storage (e.g. sets it to the master), executes a coderef and then
677 restores the original state.
683 $schema->resultset('User')->create({name=>$name});
684 my $user_rs = $schema->resultset('User')->find({name=>$name});
688 my $user_rs = $schema->storage->execute_reliably($reliably, 'John');
690 Use this when you must be certain of your database state, such as when you just
691 inserted something and need to get a resultset including it, etc.
695 sub execute_reliably {
699 $self->throw_exception('Second argument must be a coderef')
700 unless( ref $coderef eq 'CODE');
702 ## replace the current read handler for the remainder of the scope
703 local $self->{read_handler} = $self->master;
708 =head2 set_reliable_storage
710 Sets the current $schema to be 'reliable', that is all queries, both read and
711 write are sent to the master
715 sub set_reliable_storage {
717 my $schema = $self->schema;
718 my $write_handler = $self->schema->storage->write_handler;
720 $schema->storage->read_handler($write_handler);
723 =head2 set_balanced_storage
725 Sets the current $schema to be use the </balancer> for all reads, while all
726 writes are sent to the master only
730 sub set_balanced_storage {
732 my $schema = $self->schema;
733 my $balanced_handler = $self->schema->storage->balancer;
735 $schema->storage->read_handler($balanced_handler);
740 Check that the master and at least one of the replicants is connected.
747 $self->master->connected &&
748 $self->pool->connected_replicants;
751 =head2 ensure_connected
753 Make sure all the storages are connected.
757 sub ensure_connected {
759 foreach my $source ($self->all_storages) {
760 $source->ensure_connected(@_);
766 Set the limit_dialect for all existing storages
772 foreach my $source ($self->all_storages) {
773 $source->limit_dialect(@_);
775 return $self->master->limit_dialect;
780 Set the quote_char for all existing storages
786 foreach my $source ($self->all_storages) {
787 $source->quote_char(@_);
789 return $self->master->quote_char;
794 Set the name_sep for all existing storages
800 foreach my $source ($self->all_storages) {
801 $source->name_sep(@_);
803 return $self->master->name_sep;
808 Set the schema object for all existing storages
814 foreach my $source ($self->all_storages) {
815 $source->set_schema(@_);
821 set a debug flag across all storages
828 foreach my $source ($self->all_storages) {
832 return $self->master->debug;
843 return $self->master->debugobj(@_);
854 return $self->master->debugfh(@_);
865 return $self->master->debugcb(@_);
870 disconnect everything
876 foreach my $source ($self->all_storages) {
877 $source->disconnect(@_);
883 set cursor class on all storages, or return master's
888 my ($self, $cursor_class) = @_;
891 $_->cursor_class($cursor_class) for $self->all_storages;
893 $self->master->cursor_class;
898 set cursor class on all storages, or return master's, alias for L</cursor_class>
904 my ($self, $cursor_class) = @_;
907 $_->cursor($cursor_class) for $self->all_storages;
909 $self->master->cursor;
914 sets the L<DBIx::Class::Storage::DBI/unsafe> option on all storages or returns
915 master's current setting
923 $_->unsafe(@_) for $self->all_storages;
926 return $self->master->unsafe;
929 =head2 disable_sth_caching
931 sets the L<DBIx::Class::Storage::DBI/disable_sth_caching> option on all storages
932 or returns master's current setting
936 sub disable_sth_caching {
940 $_->disable_sth_caching(@_) for $self->all_storages;
943 return $self->master->disable_sth_caching;
946 =head2 lag_behind_master
948 returns the highest Replicant L<DBIx::Class::Storage::DBI/lag_behind_master>
953 sub lag_behind_master {
956 return max map $_->lag_behind_master, $self->replicants;
959 =head2 is_replicating
961 returns true if all replicants return true for
962 L<DBIx::Class::Storage::DBI/is_replicating>
969 return (grep $_->is_replicating, $self->replicants) == ($self->replicants);
972 =head2 connect_call_datetime_setup
974 calls L<DBIx::Class::Storage::DBI/connect_call_datetime_setup> for all storages
978 sub connect_call_datetime_setup {
980 $_->connect_call_datetime_setup for $self->all_storages;
985 $_->_populate_dbh for $self->all_storages;
990 $_->_connect for $self->all_storages;
995 $_->_rebless for $self->all_storages;
998 sub _determine_driver {
1000 $_->_determine_driver for $self->all_storages;
1003 sub _driver_determined {
1007 $_->_driver_determined(@_) for $self->all_storages;
1010 return $self->master->_driver_determined;
1016 $_->_init for $self->all_storages;
1019 sub _run_connection_actions {
1022 $_->_run_connection_actions for $self->all_storages;
1025 sub _do_connection_actions {
1029 $_->_do_connection_actions(@_) for $self->all_storages;
1033 sub connect_call_do_sql {
1035 $_->connect_call_do_sql(@_) for $self->all_storages;
1038 sub disconnect_call_do_sql {
1040 $_->disconnect_call_do_sql(@_) for $self->all_storages;
1043 sub _seems_connected {
1046 return min map $_->_seems_connected, $self->all_storages;
1052 return min map $_->_ping, $self->all_storages;
1057 Due to the fact that replicants can lag behind a master, you must take care to
1058 make sure you use one of the methods to force read queries to a master should
1059 you need realtime data integrity. For example, if you insert a row, and then
1060 immediately re-read it from the database (say, by doing
1061 L<< $result->discard_changes|DBIx::Class::Row/discard_changes >>)
1062 or you insert a row and then immediately build a query that expects that row
1063 to be an item, you should force the master to handle reads. Otherwise, due to
1064 the lag, there is no certainty your data will be in the expected state.
1066 For data integrity, all transactions automatically use the master storage for
1067 all read and write queries. Using a transaction is the preferred and recommended
1068 method to force the master to handle all read queries.
1070 Otherwise, you can force a single query to use the master with the 'force_pool'
1073 my $result = $resultset->search(undef, {force_pool=>'master'})->find($pk);
1075 This attribute will safely be ignored by non replicated storages, so you can use
1076 the same code for both types of systems.
1078 Lastly, you can use the L</execute_reliably> method, which works very much like
1081 For debugging, you can turn replication on/off with the methods L</set_reliable_storage>
1082 and L</set_balanced_storage>, however this operates at a global level and is not
1083 suitable if you have a shared Schema object being used by multiple processes,
1084 such as on a web application server. You can get around this limitation by
1085 using the Schema clone method.
1087 my $new_schema = $schema->clone;
1088 $new_schema->set_reliable_storage;
1090 ## $new_schema will use only the Master storage for all reads/writes while
1091 ## the $schema object will use replicated storage.
1093 =head1 FURTHER QUESTIONS?
1095 Check the list of L<additional DBIC resources|DBIx::Class/GETTING HELP/SUPPORT>.
1097 =head1 COPYRIGHT AND LICENSE
1099 This module is free software L<copyright|DBIx::Class/COPYRIGHT AND LICENSE>
1100 by the L<DBIx::Class (DBIC) authors|DBIx::Class/AUTHORS>. You can
1101 redistribute it and/or modify it under the same terms as the
1102 L<DBIx::Class library|DBIx::Class/COPYRIGHT AND LICENSE>.
1106 __PACKAGE__->meta->make_immutable;