X-Git-Url: http://git.shadowcat.co.uk/gitweb/gitweb.cgi?a=blobdiff_plain;f=lib%2FDBIx%2FClass%2FStorage%2FDBI%2FReplicated.pm;h=6302bf1ebac360671bb02075f28ab3ebcd8e66c3;hb=212cc5c25c31b2ec3ff4b4e20283321617db79e6;hp=df0734faa21e2831d6c0ac9b9de2065a0cb22bc6;hpb=955a6df69eb864ba348f87f5a06a8b7a68ffc76c;p=dbsrgits%2FDBIx-Class.git diff --git a/lib/DBIx/Class/Storage/DBI/Replicated.pm b/lib/DBIx/Class/Storage/DBI/Replicated.pm index df0734f..6302bf1 100644 --- a/lib/DBIx/Class/Storage/DBI/Replicated.pm +++ b/lib/DBIx/Class/Storage/DBI/Replicated.pm @@ -1,16 +1,46 @@ package DBIx::Class::Storage::DBI::Replicated; +BEGIN { + use Carp::Clan qw/^DBIx::Class/; + + ## Modules required for Replication support not required for general DBIC + ## use, so we explicitly test for these. + + my %replication_required = ( + 'Moose' => '0.77', + 'MooseX::AttributeHelpers' => '0.12', + 'MooseX::Types' => '0.10', + 'namespace::clean' => '0.11', + 'Hash::Merge' => '0.11' + ); + + my @didnt_load; + + for my $module (keys %replication_required) { + eval "use $module $replication_required{$module}"; + push @didnt_load, "$module $replication_required{$module}" + if $@; + } + + croak("@{[ join ', ', @didnt_load ]} are missing and are required for Replication") + if @didnt_load; +} + use Moose; use DBIx::Class::Storage::DBI; use DBIx::Class::Storage::DBI::Replicated::Pool; use DBIx::Class::Storage::DBI::Replicated::Balancer; -use Scalar::Util qw(blessed); +use DBIx::Class::Storage::DBI::Replicated::Types 'BalancerClassNamePart'; +use MooseX::Types::Moose qw/ClassName HashRef Object/; +use Scalar::Util 'reftype'; +use Carp::Clan qw/^DBIx::Class/; +use Hash::Merge 'merge'; -extends 'DBIx::Class::Storage::DBI', 'Moose::Object'; +use namespace::clean -except => 'meta'; =head1 NAME -DBIx::Class::Storage::DBI::Replicated - ALPHA Replicated database support +DBIx::Class::Storage::DBI::Replicated - BETA Replicated database support =head1 SYNOPSIS @@ -18,24 +48,52 @@ The Following example shows how to change an existing $schema to a replicated storage type, add some replicated (readonly) databases, and perform reporting tasks. - ## Change storage_type in your schema class - $schema->storage_type( '::DBI::Replicated' ); - - ## Add some slaves. Basically this is an array of arrayrefs, where each - ## arrayref is database connect information - - $schema->storage->connect_replicants( - [$dsn1, $user, $pass, \%opts], - [$dsn1, $user, $pass, \%opts], - [$dsn1, $user, $pass, \%opts], - ); - +You should set the 'storage_type attribute to a replicated type. You should +also defined you arguments, such as which balancer you want and any arguments +that the Pool object should get. + + $schema->storage_type( ['::DBI::Replicated', {balancer=>'::Random'}] ); + +Next, you need to add in the Replicants. Basically this is an array of +arrayrefs, where each arrayref is database connect information. Think of these +arguments as what you'd pass to the 'normal' $schema->connect method. + + $schema->storage->connect_replicants( + [$dsn1, $user, $pass, \%opts], + [$dsn2, $user, $pass, \%opts], + [$dsn3, $user, $pass, \%opts], + ); + +Now, just use the $schema as you normally would. Automatically all reads will +be delegated to the replicants, while writes to the master. + + $schema->resultset('Source')->search({name=>'etc'}); + +You can force a given query to use a particular storage using the search +attribute 'force_pool'. For example: + + my $RS = $schema->resultset('Source')->search(undef, {force_pool=>'master'}); + +Now $RS will force everything (both reads and writes) to use whatever was setup +as the master storage. 'master' is hardcoded to always point to the Master, +but you can also use any Replicant name. Please see: +L and the replicants attribute for more. + +Also see transactions and L for alternative ways to +force read traffic to the master. In general, you should wrap your statements +in a transaction when you are reading and writing to the same tables at the +same time, since your replicants will often lag a bit behind the master. + +See L for more help and +walkthroughs. + =head1 DESCRIPTION -Warning: This class is marked ALPHA. We are using this in development and have -some basic test coverage but the code hasn't yet been stressed by a variety -of databases. Individual DB's may have quirks we are not aware of. Please -use this in development and pass along your experiences/bug fixes. +Warning: This class is marked BETA. This has been running a production +website using MySQL native replication as its backend and we have some decent +test coverage but the code hasn't yet been stressed by a variety of databases. +Individual DB's may have quirks we are not aware of. Please use this in first +development and pass along your experiences/bug fixes. This class implements replicated data store for DBI. Currently you can define one master and numerous slave database connections. All write-type queries @@ -43,89 +101,51 @@ one master and numerous slave database connections. All write-type queries database, all read-type queries (SELECTs) go to the slave database. Basically, any method request that L would normally -handle gets delegated to one of the two attributes: L or to -L. Additionally, some methods need to be distributed +handle gets delegated to one of the two attributes: L or to +L. Additionally, some methods need to be distributed to all existing storages. This way our storage class is a drop in replacement for L. Read traffic is spread across the replicants (slaves) occuring to a user selected algorithm. The default algorithm is random weighted. -TODO more details about the algorithm. +=head1 NOTES -=head1 ATTRIBUTES +The consistancy betweeen master and replicants is database specific. The Pool +gives you a method to validate it's replicants, removing and replacing them +when they fail/pass predefined criteria. Please make careful use of the ways +to force a query to run against Master when needed. -This class defines the following attributes. +=head1 REQUIREMENTS -=head2 master - -The master defines the canonical state for a pool of connected databases. All -the replicants are expected to match this databases state. Thus, in a classic -Master / Slaves distributed system, all the slaves are expected to replicate -the Master's state as quick as possible. This is the only database in the -pool of databases that is allowed to handle write traffic. - -=cut - -has 'master' => ( - is=> 'ro', - isa=>'DBIx::Class::Storage::DBI', - lazy_build=>1, - handles=>[qw/ - on_connect_do - on_disconnect_do - connect_info - throw_exception - sql_maker - sqlt_type - create_ddl_dir - deployment_statements - datetime_parser - datetime_parser_type - last_insert_id - insert - insert_bulk - update - delete - dbh - txn_do - txn_commit - txn_rollback - sth - deploy - schema - /], -); +Replicated Storage has additional requirements not currently part of L + Moose => 0.77 + MooseX::AttributeHelpers => 0.12 + MooseX::Types => 0.10 + namespace::clean => 0.11 + Hash::Merge => 0.11 + +You will need to install these modules manually via CPAN or make them part of the +Makefile for your distribution. -=head2 current_replicant +=head1 ATTRIBUTES -Replicant storages (slaves) handle all read only traffic. The assumption is -that your database will become readbound well before it becomes write bound -and that being able to spread your read only traffic around to multiple -databases is going to help you to scale traffic. +This class defines the following attributes. -This attribute returns the next slave to handle a read request. Your L -attribute has methods to help you shuffle through all the available replicants -via it's balancer object. +=head2 schema -We split the reader/writer to make it easier to selectively override how the -replicant is altered. +The underlying L object this storage is attaching =cut -has 'current_replicant' => ( - is=> 'rw', - isa=>'DBIx::Class::Storage::DBI', - lazy_build=>1, - handles=>[qw/ - select - select_single - columns_info_for - /], +has 'schema' => ( + is=>'rw', + isa=>'DBIx::Class::Schema', + weak_ref=>1, + required=>1, ); - =head2 pool_type Contains the classname which will instantiate the L object. Defaults @@ -134,14 +154,26 @@ to: L. =cut has 'pool_type' => ( - is=>'ro', - isa=>'ClassName', - required=>1, - lazy=>1, - default=>'DBIx::Class::Storage::DBI::Replicated::Pool', - handles=>{ - 'create_pool' => 'new', - }, + is=>'rw', + isa=>ClassName, + default=>'DBIx::Class::Storage::DBI::Replicated::Pool', + handles=>{ + 'create_pool' => 'new', + }, +); + +=head2 pool_args + +Contains a hashref of initialized information to pass to the Balancer object. +See L for available arguments. + +=cut + +has 'pool_args' => ( + is=>'rw', + isa=>HashRef, + lazy=>1, + default=>sub { {} }, ); @@ -153,16 +185,30 @@ choose how to spread the query load across each replicant in the pool. =cut has 'balancer_type' => ( - is=>'ro', - isa=>'ClassName', - required=>1, - lazy=>1, - default=>'DBIx::Class::Storage::DBI::Replicated::Balancer', - handles=>{ - 'create_balancer' => 'new', - }, + is=>'rw', + isa=>BalancerClassNamePart, + coerce=>1, + required=>1, + default=> 'DBIx::Class::Storage::DBI::Replicated::Balancer::First', + handles=>{ + 'create_balancer' => 'new', + }, ); +=head2 balancer_args + +Contains a hashref of initialized information to pass to the Balancer object. +See L for available arguments. + +=cut + +has 'balancer_args' => ( + is=>'rw', + isa=>HashRef, + lazy=>1, + required=>1, + default=>sub { {} }, +); =head2 pool @@ -172,19 +218,16 @@ container class for one or more replicated databases. =cut has 'pool' => ( - is=>'ro', - isa=>'DBIx::Class::Storage::DBI::Replicated::Pool', - lazy_build=>1, - handles=>[qw/ - replicants - has_replicants - connect_replicants - num_replicants - delete_replicant - /], + is=>'ro', + isa=>'DBIx::Class::Storage::DBI::Replicated::Pool', + lazy_build=>1, + handles=>[qw/ + connect_replicants + replicants + has_replicants + /], ); - =head2 balancer Is a or derived class. This @@ -193,39 +236,201 @@ is a class that takes a pool () =cut has 'balancer' => ( - is=>'ro', - isa=>'DBIx::Class::Storage::DBI::Replicated::Balancer', - lazy_build=>1, - handles=>[qw/next_storage/], + is=>'rw', + isa=>'DBIx::Class::Storage::DBI::Replicated::Balancer', + lazy_build=>1, + handles=>[qw/auto_validate_every/], ); +=head2 master + +The master defines the canonical state for a pool of connected databases. All +the replicants are expected to match this databases state. Thus, in a classic +Master / Slaves distributed system, all the slaves are expected to replicate +the Master's state as quick as possible. This is the only database in the +pool of databases that is allowed to handle write traffic. + +=cut + +has 'master' => ( + is=> 'ro', + isa=>'DBIx::Class::Storage::DBI', + lazy_build=>1, +); + +=head1 ATTRIBUTES IMPLEMENTING THE DBIx::Storage::DBI INTERFACE + +The following methods are delegated all the methods required for the +L interface. + +=head2 read_handler + +Defines an object that implements the read side of L. + +=cut + +has 'read_handler' => ( + is=>'rw', + isa=>Object, + lazy_build=>1, + handles=>[qw/ + select + select_single + columns_info_for + /], +); + +=head2 write_handler + +Defines an object that implements the write side of L. + +=cut + +has 'write_handler' => ( + is=>'ro', + isa=>Object, + lazy_build=>1, + handles=>[qw/ + on_connect_do + on_disconnect_do + connect_info + throw_exception + sql_maker + sqlt_type + create_ddl_dir + deployment_statements + datetime_parser + datetime_parser_type + build_datetime_parser + last_insert_id + insert + insert_bulk + update + delete + dbh + txn_begin + txn_do + txn_commit + txn_rollback + txn_scope_guard + sth + deploy + with_deferred_fk_checks + dbh_do + reload_row + with_deferred_fk_checks + _prep_for_execute + + backup + is_datatype_numeric + _count_select + _subq_count_select + _subq_update_delete + svp_rollback + svp_begin + svp_release + /], +); + +has _master_connect_info_opts => + (is => 'rw', isa => HashRef, default => sub { {} }); + +=head2 around: connect_info + +Preserve master's C options (for merging with replicants.) +Also set any Replicated related options from connect_info, such as +C, C, C and C. + +=cut + +around connect_info => sub { + my ($next, $self, $info, @extra) = @_; + + my $wantarray = wantarray; + + my %opts; + for my $arg (@$info) { + next unless (reftype($arg)||'') eq 'HASH'; + %opts = %{ merge($arg, \%opts) }; + } + delete $opts{dsn}; + + if (@opts{qw/pool_type pool_args/}) { + $self->pool_type(delete $opts{pool_type}) + if $opts{pool_type}; + + $self->pool_args( + merge((delete $opts{pool_args} || {}), $self->pool_args) + ); + + $self->pool($self->_build_pool) + if $self->pool; + } + + if (@opts{qw/balancer_type balancer_args/}) { + $self->balancer_type(delete $opts{balancer_type}) + if $opts{balancer_type}; + + $self->balancer_args( + merge((delete $opts{balancer_args} || {}), $self->balancer_args) + ); + + $self->balancer($self->_build_balancer) + if $self->balancer; + } + + $self->_master_connect_info_opts(\%opts); + + my (@res, $res); + if ($wantarray) { + @res = $self->$next($info, @extra); + } else { + $res = $self->$next($info, @extra); + } + + # Make sure master is blessed into the correct class and apply role to it. + my $master = $self->master; + $master->_determine_driver; + Moose::Meta::Class->initialize(ref $master); + DBIx::Class::Storage::DBI::Replicated::WithDSN->meta->apply($master); + + $wantarray ? @res : $res; +}; + =head1 METHODS This class defines the following methods. -=head2 _build_master +=head2 BUILDARGS -Lazy builder for the L attribute. +L when instantiating it's storage passed itself as the +first argument. So we need to massage the arguments a bit so that all the +bits get put into the correct places. =cut -sub _build_master { - DBIx::Class::Storage::DBI->new; +sub BUILDARGS { + my ($class, $schema, $storage_type_args, @args) = @_; + + return { + schema=>$schema, + %$storage_type_args, + @args + } } +=head2 _build_master -=head2 _build_current_replicant - -Lazy builder for the L attribute. +Lazy builder for the L attribute. =cut -sub _build_current_replicant { - my $self = shift @_; - $self->next_storage($self->pool); +sub _build_master { + my $self = shift @_; + my $master = DBIx::Class::Storage::DBI->new($self->schema); + $master } - =head2 _build_pool Lazy builder for the L attribute. @@ -233,63 +438,94 @@ Lazy builder for the L attribute. =cut sub _build_pool { - my $self = shift @_; - $self->create_pool; + my $self = shift @_; + $self->create_pool(%{$self->pool_args}); } - =head2 _build_balancer -Lazy builder for the L attribute. +Lazy builder for the L attribute. This takes a Pool object so that +the balancer knows which pool it's balancing. =cut sub _build_balancer { - my $self = shift @_; - $self->create_balancer; + my $self = shift @_; + $self->create_balancer( + pool=>$self->pool, + master=>$self->master, + %{$self->balancer_args}, + ); } +=head2 _build_write_handler -=head2 around: create_replicants - -All calls to create_replicants needs to have an existing $schema tacked onto -top of the args +Lazy builder for the L attribute. The default is to set this to +the L. =cut -around 'connect_replicants' => sub { - my ($method, $self, @args) = @_; - $self->$method($self->schema, @args); -}; +sub _build_write_handler { + return shift->master; +} + +=head2 _build_read_handler + +Lazy builder for the L attribute. The default is to set this to +the L. + +=cut +sub _build_read_handler { + return shift->balancer; +} -=head2 after: select, select_single, columns_info_for +=head2 around: connect_replicants -Advice on the current_replicant_storage attribute. Each time we use a replicant -we need to change it via the storage pool algorithm. That way we are spreading -the load evenly (hopefully) across existing capacity. +All calls to connect_replicants needs to have an existing $schema tacked onto +top of the args, since L needs it, and any C +options merged with the master, with replicant opts having higher priority. =cut -after 'select' => sub { - my $self = shift @_; - my $next_replicant = $self->next_storage($self->pool); +around connect_replicants => sub { + my ($next, $self, @args) = @_; - $self->current_replicant($next_replicant); -}; + for my $r (@args) { + $r = [ $r ] unless reftype $r eq 'ARRAY'; -after 'select_single' => sub { - my $self = shift @_; - my $next_replicant = $self->next_storage($self->pool); + croak "coderef replicant connect_info not supported" + if ref $r->[0] && reftype $r->[0] eq 'CODE'; - $self->current_replicant($next_replicant); -}; +# any connect_info options? + my $i = 0; + $i++ while $i < @$r && (reftype($r->[$i])||'') ne 'HASH'; -after 'columns_info_for' => sub { - my $self = shift @_; - my $next_replicant = $self->next_storage($self->pool); +# make one if none + $r->[$i] = {} unless $r->[$i]; - $self->current_replicant($next_replicant); +# merge if two hashes + my @hashes = @$r[$i .. $#{$r}]; + + croak "invalid connect_info options" + if (grep { reftype($_) eq 'HASH' } @hashes) != @hashes; + + croak "too many hashrefs in connect_info" + if @hashes > 2; + + my %opts = %{ merge(reverse @hashes) }; + +# delete them + splice @$r, $i+1, ($#{$r} - $i), (); + +# merge with master + %opts = %{ merge(\%opts, $self->_master_connect_info_opts) }; + +# update + $r->[$i] = \%opts; + } + + $self->$next($self->schema, @args); }; =head2 all_storages @@ -301,14 +537,107 @@ replicants. =cut sub all_storages { - my $self = shift @_; - - return grep {defined $_ && blessed $_} ( - $self->master, - $self->replicants, - ); + my $self = shift @_; + return grep {defined $_ && blessed $_} ( + $self->master, + values %{ $self->replicants }, + ); } +=head2 execute_reliably ($coderef, ?@args) + +Given a coderef, saves the current state of the L, forces it to +use reliable storage (ie sets it to the master), executes a coderef and then +restores the original state. + +Example: + + my $reliably = sub { + my $name = shift @_; + $schema->resultset('User')->create({name=>$name}); + my $user_rs = $schema->resultset('User')->find({name=>$name}); + return $user_rs; + }; + + my $user_rs = $schema->storage->execute_reliably($reliably, 'John'); + +Use this when you must be certain of your database state, such as when you just +inserted something and need to get a resultset including it, etc. + +=cut + +sub execute_reliably { + my ($self, $coderef, @args) = @_; + + unless( ref $coderef eq 'CODE') { + $self->throw_exception('Second argument must be a coderef'); + } + + ##Get copy of master storage + my $master = $self->master; + + ##Get whatever the current read hander is + my $current = $self->read_handler; + + ##Set the read handler to master + $self->read_handler($master); + + ## do whatever the caller needs + my @result; + my $want_array = wantarray; + + eval { + if($want_array) { + @result = $coderef->(@args); + } elsif(defined $want_array) { + ($result[0]) = ($coderef->(@args)); + } else { + $coderef->(@args); + } + }; + + ##Reset to the original state + $self->read_handler($current); + + ##Exception testing has to come last, otherwise you might leave the + ##read_handler set to master. + + if($@) { + $self->throw_exception("coderef returned an error: $@"); + } else { + return $want_array ? @result : $result[0]; + } +} + +=head2 set_reliable_storage + +Sets the current $schema to be 'reliable', that is all queries, both read and +write are sent to the master + +=cut + +sub set_reliable_storage { + my $self = shift @_; + my $schema = $self->schema; + my $write_handler = $self->schema->storage->write_handler; + + $schema->storage->read_handler($write_handler); +} + +=head2 set_balanced_storage + +Sets the current $schema to be use the for all reads, while all +writea are sent to the master only + +=cut + +sub set_balanced_storage { + my $self = shift @_; + my $schema = $self->schema; + my $balanced_handler = $self->schema->storage->balancer; + + $schema->storage->read_handler($balanced_handler); +} =head2 connected @@ -317,14 +646,12 @@ Check that the master and at least one of the replicants is connected. =cut sub connected { - my $self = shift @_; - - return - $self->master->connected && - $self->pool->connected_replicants; + my $self = shift @_; + return + $self->master->connected && + $self->pool->connected_replicants; } - =head2 ensure_connected Make sure all the storages are connected. @@ -332,13 +659,12 @@ Make sure all the storages are connected. =cut sub ensure_connected { - my $self = shift @_; - foreach my $source ($self->all_storages) { - $source->ensure_connected(@_); - } + my $self = shift @_; + foreach my $source ($self->all_storages) { + $source->ensure_connected(@_); + } } - =head2 limit_dialect Set the limit_dialect for all existing storages @@ -346,13 +672,13 @@ Set the limit_dialect for all existing storages =cut sub limit_dialect { - my $self = shift @_; - foreach my $source ($self->all_storages) { - $source->limit_dialect(@_); - } + my $self = shift @_; + foreach my $source ($self->all_storages) { + $source->limit_dialect(@_); + } + return $self->master->quote_char; } - =head2 quote_char Set the quote_char for all existing storages @@ -360,13 +686,13 @@ Set the quote_char for all existing storages =cut sub quote_char { - my $self = shift @_; - foreach my $source ($self->all_storages) { - $source->quote_char(@_); - } + my $self = shift @_; + foreach my $source ($self->all_storages) { + $source->quote_char(@_); + } + return $self->master->quote_char; } - =head2 name_sep Set the name_sep for all existing storages @@ -374,13 +700,13 @@ Set the name_sep for all existing storages =cut sub name_sep { - my $self = shift @_; - foreach my $source ($self->all_storages) { - $source->name_sep(@_); - } + my $self = shift @_; + foreach my $source ($self->all_storages) { + $source->name_sep(@_); + } + return $self->master->name_sep; } - =head2 set_schema Set the schema object for all existing storages @@ -388,13 +714,12 @@ Set the schema object for all existing storages =cut sub set_schema { - my $self = shift @_; - foreach my $source ($self->all_storages) { - $source->set_schema(@_); - } + my $self = shift @_; + foreach my $source ($self->all_storages) { + $source->set_schema(@_); + } } - =head2 debug set a debug flag across all storages @@ -402,13 +727,15 @@ set a debug flag across all storages =cut sub debug { - my $self = shift @_; + my $self = shift @_; + if(@_) { foreach my $source ($self->all_storages) { - $source->debug(@_); - } + $source->debug(@_); + } + } + return $self->master->debug; } - =head2 debugobj set a debug object across all storages @@ -416,13 +743,15 @@ set a debug object across all storages =cut sub debugobj { - my $self = shift @_; + my $self = shift @_; + if(@_) { foreach my $source ($self->all_storages) { - $source->debugobj(@_); - } + $source->debugobj(@_); + } + } + return $self->master->debugobj; } - =head2 debugfh set a debugfh object across all storages @@ -430,13 +759,15 @@ set a debugfh object across all storages =cut sub debugfh { - my $self = shift @_; + my $self = shift @_; + if(@_) { foreach my $source ($self->all_storages) { - $source->debugfh(@_); - } + $source->debugfh(@_); + } + } + return $self->master->debugfh; } - =head2 debugcb set a debug callback across all storages @@ -444,13 +775,15 @@ set a debug callback across all storages =cut sub debugcb { - my $self = shift @_; + my $self = shift @_; + if(@_) { foreach my $source ($self->all_storages) { - $source->debugcb(@_); - } + $source->debugcb(@_); + } + } + return $self->master->debugcb; } - =head2 disconnect disconnect everything @@ -458,305 +791,72 @@ disconnect everything =cut sub disconnect { - my $self = shift @_; - foreach my $source ($self->all_storages) { - $source->disconnect(@_); - } + my $self = shift @_; + foreach my $source ($self->all_storages) { + $source->disconnect(@_); + } } +=head2 cursor_class -=head2 DESTROY - -Make sure we pass destroy events down to the storage handlers - -=cut - -sub DESTROY { - my $self = shift; - ## TODO, maybe we can just leave this alone ??? -} - - -=head1 AUTHOR - -Norbert Csongrádi - -Peter Siklósi - -John Napiorkowski - -=head1 LICENSE - -You may distribute this code under the same terms as Perl itself. - -=cut - -1; - -__END__ - -use strict; -use warnings; - -use DBIx::Class::Storage::DBI; -use DBD::Multi; - -use base qw/Class::Accessor::Fast/; - -__PACKAGE__->mk_accessors( qw/read_source write_source/ ); - -=head1 NAME - -DBIx::Class::Storage::DBI::Replicated - ALPHA Replicated database support - -=head1 SYNOPSIS - -The Following example shows how to change an existing $schema to a replicated -storage type and update it's connection information to contain a master DSN and -an array of slaves. - - ## Change storage_type in your schema class - $schema->storage_type( '::DBI::Replicated' ); - - ## Set your connection. - $schema->connect( - $dsn, $user, $password, { - AutoCommit => 1, - ## Other standard DBI connection or DBD custom attributes added as - ## usual. Additionally, we have two custom attributes for defining - ## slave information and controlling how the underlying DBD::Multi - connect_replicants => [ - ## Define each slave like a 'normal' DBI connection, but you add - ## in a DBD::Multi custom attribute to define how the slave is - ## prioritized. Please see DBD::Multi for more. - [$slave1dsn, $user, $password, {%slave1opts}], - [$slave2dsn, $user, $password, {%slave2opts}], - [$slave3dsn, $user, $password, {%slave3opts}], - ], - }, - ); - - ## Now, just use the schema as normal - $schema->resultset('Table')->find(< unique >); ## Reads will use slaves - $schema->resultset('Table')->create(\%info); ## Writes will use master - -=head1 DESCRIPTION - -Warning: This class is marked ALPHA. We are using this in development and have -some basic test coverage but the code hasn't yet been stressed by a variety -of databases. Individual DB's may have quirks we are not aware of. Please -use this in development and pass along your experiences/bug fixes. - -This class implements replicated data store for DBI. Currently you can define -one master and numerous slave database connections. All write-type queries -(INSERT, UPDATE, DELETE and even LAST_INSERT_ID) are routed to master -database, all read-type queries (SELECTs) go to the slave database. - -For every slave database you can define a priority value, which controls data -source usage pattern. It uses L, so first the lower priority data -sources used (if they have the same priority, the are used randomized), than -if all low priority data sources fail, higher ones tried in order. - -=head1 CONFIGURATION - -Please see L for most configuration information. +set cursor class on all storages, or return master's =cut -sub new { - my $proto = shift; - my $class = ref( $proto ) || $proto; - my $self = {}; - - bless( $self, $class ); - - $self->write_source( DBIx::Class::Storage::DBI->new ); - $self->read_source( DBIx::Class::Storage::DBI->new ); +sub cursor_class { + my ($self, $cursor_class) = @_; - return $self; + if ($cursor_class) { + $_->cursor_class($cursor_class) for $self->all_storages; + } + $self->master->cursor_class; } - -sub all_sources { - my $self = shift; - - my @sources = ($self->read_source, $self->write_source); - - return wantarray ? @sources : \@sources; -} - -sub _connect_info { - my $self = shift; - my $master = $self->write_source->_connect_info; - $master->[-1]->{slave_connect_info} = $self->read_source->_connect_info; - return $master; -} - -sub connect_info { - my ($self, $source_info) = @_; - - ## if there is no $source_info, treat this sub like an accessor - return $self->_connect_info - if !$source_info; - - ## Alright, let's conect the master - $self->write_source->connect_info($source_info); - ## Now, build and then connect the Slaves - my @slaves_connect_info = @{$source_info->[-1]->{slaves_connect_info}}; - my $dbd_multi_config = ref $slaves_connect_info[-1] eq 'HASH' - ? pop @slaves_connect_info : {}; - - ## We need to do this since SQL::Abstract::Limit can't guess what DBD::Multi is - $dbd_multi_config->{limit_dialect} = $self->write_source->sql_maker->limit_dialect - unless defined $dbd_multi_config->{limit_dialect}; - - @slaves_connect_info = map { - ## if the first element in the arrayhash is a ref, make that the value - my $db = ref $_->[0] ? $_->[0] : $_; - my $priority = $_->[-1]->{priority} || 10; ## default priority is 10 - $priority => $db; - } @slaves_connect_info; - - $self->read_source->connect_info([ - 'dbi:Multi:', undef, undef, { - dsns => [@slaves_connect_info], - %$dbd_multi_config, - }, - ]); - - ## Return the formated connection information - return $self->_connect_info; -} - -sub select { - shift->read_source->select( @_ ); -} -sub select_single { - shift->read_source->select_single( @_ ); -} -sub throw_exception { - shift->read_source->throw_exception( @_ ); -} -sub sql_maker { - shift->read_source->sql_maker( @_ ); -} -sub columns_info_for { - shift->read_source->columns_info_for( @_ ); -} -sub sqlt_type { - shift->read_source->sqlt_type( @_ ); -} -sub create_ddl_dir { - shift->read_source->create_ddl_dir( @_ ); -} -sub deployment_statements { - shift->read_source->deployment_statements( @_ ); -} -sub datetime_parser { - shift->read_source->datetime_parser( @_ ); -} -sub datetime_parser_type { - shift->read_source->datetime_parser_type( @_ ); -} -sub build_datetime_parser { - shift->read_source->build_datetime_parser( @_ ); -} - -sub limit_dialect { $_->limit_dialect( @_ ) for( shift->all_sources ) } -sub quote_char { $_->quote_char( @_ ) for( shift->all_sources ) } -sub name_sep { $_->quote_char( @_ ) for( shift->all_sources ) } -sub disconnect { $_->disconnect( @_ ) for( shift->all_sources ) } -sub set_schema { $_->set_schema( @_ ) for( shift->all_sources ) } +=head1 GOTCHAS -sub DESTROY { - my $self = shift; +Due to the fact that replicants can lag behind a master, you must take care to +make sure you use one of the methods to force read queries to a master should +you need realtime data integrity. For example, if you insert a row, and then +immediately re-read it from the database (say, by doing $row->discard_changes) +or you insert a row and then immediately build a query that expects that row +to be an item, you should force the master to handle reads. Otherwise, due to +the lag, there is no certainty your data will be in the expected state. - undef $self->{write_source}; - undef $self->{read_sources}; -} - -sub last_insert_id { - shift->write_source->last_insert_id( @_ ); -} -sub insert { - shift->write_source->insert( @_ ); -} -sub update { - shift->write_source->update( @_ ); -} -sub update_all { - shift->write_source->update_all( @_ ); -} -sub delete { - shift->write_source->delete( @_ ); -} -sub delete_all { - shift->write_source->delete_all( @_ ); -} -sub create { - shift->write_source->create( @_ ); -} -sub find_or_create { - shift->write_source->find_or_create( @_ ); -} -sub update_or_create { - shift->write_source->update_or_create( @_ ); -} -sub connected { - shift->write_source->connected( @_ ); -} -sub ensure_connected { - shift->write_source->ensure_connected( @_ ); -} -sub dbh { - shift->write_source->dbh( @_ ); -} -sub txn_do { - shift->write_source->txn_do( @_ ); -} -sub txn_commit { - shift->write_source->txn_commit( @_ ); -} -sub txn_rollback { - shift->write_source->txn_rollback( @_ ); -} -sub sth { - shift->write_source->sth( @_ ); -} -sub deploy { - shift->write_source->deploy( @_ ); -} -sub _prep_for_execute { - shift->write_source->_prep_for_execute(@_); -} +For data integrity, all transactions automatically use the master storage for +all read and write queries. Using a transaction is the preferred and recommended +method to force the master to handle all read queries. -sub debugobj { - shift->write_source->debugobj(@_); -} -sub debug { - shift->write_source->debug(@_); -} +Otherwise, you can force a single query to use the master with the 'force_pool' +attribute: -sub debugfh { shift->_not_supported( 'debugfh' ) }; -sub debugcb { shift->_not_supported( 'debugcb' ) }; + my $row = $resultset->search(undef, {force_pool=>'master'})->find($pk); -sub _not_supported { - my( $self, $method ) = @_; +This attribute will safely be ignore by non replicated storages, so you can use +the same code for both types of systems. - die "This Storage does not support $method method."; -} +Lastly, you can use the L method, which works very much like +a transaction. -=head1 SEE ALSO +For debugging, you can turn replication on/off with the methods L +and L, however this operates at a global level and is not +suitable if you have a shared Schema object being used by multiple processes, +such as on a web application server. You can get around this limitation by +using the Schema clone method. -L, L, L + my $new_schema = $schema->clone; + $new_schema->set_reliable_storage; + + ## $new_schema will use only the Master storage for all reads/writes while + ## the $schema object will use replicated storage. =head1 AUTHOR -Norbert Csongrádi + John Napiorkowski -Peter Siklósi +Based on code originated by: -John Napiorkowski + Norbert Csongrádi + Peter Siklósi =head1 LICENSE @@ -764,4 +864,6 @@ You may distribute this code under the same terms as Perl itself. =cut +__PACKAGE__->meta->make_immutable; + 1;