View on
MetaCPAN
Toru Yamaguchi > DBIx-DBHResolver-0.13 > DBIx::DBHResolver

Download:
DBIx-DBHResolver-0.13.tar.gz

Dependencies

Annotate this POD

CPAN RT

New  2
Open  0
View/Report Bugs
Module Version: 0.13   Source   Latest Release: DBIx-DBHResolver-0.17

NAME ^

DBIx::DBHResolver - Resolve database connection on the environment has many database servers.

SYNOPSIS ^

  use DBIx::DBHResolver;

  my $r = DBIx::DBHResolver->new;
  $r->config(+{
    connect_info => +{
      main_master => +{
        dsn => 'dbi:mysql:dbname=main;host=localhost',
        user => 'master_user', password => '',
        attrs => +{ RaiseError => 1, AutoCommit => 0, },
      },
      main_slave => +{
        dsn => 'dbi:mysql:dbname=main;host=localhost',
        user => 'slave_user', password => '',
        attrs => +{ RaiseError => 1, AutoCommit => 1, },
      }
    },
  });

  my $dbh_master = $r->connect('main_master');
  $dbh_master->do( 'UPDATE people SET ...', undef, ... );

  my $dbh_slave = $r->connect('main_slave');
  my $people = $dbh_slave->selectrow_hashref( 'SELECT * FROM people WHERE id = ?', undef, 20 );

DESCRIPTION ^

DBIx::DBHResolver resolves database connection on the environment has many database servers. The resolution algorithm is extensible and pluggable, because of this you can make custom strategy module easily.

This module can retrieve DBI's database handle object or connection information (data source, user, credential...) by labeled name and treat same cluster consists many nodes as one labeled name, choose fetching strategy.

DBIx::DBHResolver is able to use as instance or static class.

USING STRATEGY, MAKING CUSTOM STRATEGY

See DBIx::DBHResolver::Strategy::Key.

METHODS ^

new()

Create DBIx::DBHResolver instance.

load( $yaml_file_path )

Load config file formatted yaml.

config( \%config )

Load config. Example config (perl hash reference format):

  +{
    clusters => +{
      diary_master => [qw/diary001_master diary002_master/],
      people_master => [qw/people001_master people002_master people003_master people004_master/]
    },
    connect_info => +{
      diary001_master => +{
        dsn => 'dbi:driverName:...',
        user => 'root', password => '', attrs => +{},
      },
      diary002_master => +{ ... },
      ...
    },
  }

connect( $cluster_or_node, $args )

Retrieve database handle. If $args is scalar or array reference, then $args is treated sharding key. If $args is hash reference, then see below.

strategy

Optional parameter. Specify suffix of strategy module name. Default strategy module is prefixed 'DBIx::DBHResolver::Strategy::'. If you want to make custom strategy that is not started of 'DBIx::DBHResolver::Strategy::', then add prefix '+' at the beginning of the module name, such as '+MyApp::Strategy::Custom'.

key

Optional parameter. Strategy module uses hint choosing node.

connect_cached($cluster_or_node, $args)

Retrieve database handle from own cache, if not exists cache then using DBI::connect(). $args is same as connect().

connect_info($cluster_or_node, $args)

Retrieve connection info as HASHREF. $args is same as connect().

resolve($cluster_or_node, $args)

Return resolved node name. $args is same as connect.

resolve_node_keys($cluster_or_node, $keys, $args)

Return hash resolved node and keys. $args is same as connect

  use DBIx::DBHResolver;

  my $resolver = DBIx::DBHResolver->new;
  $resolver->config(+{
    clusters => +{
      MASTER => +{
        nodes => [qw/MASTER001 MASTER002 MASTER003/],
        strategy => 'Key',
      }
    },
    connect_info => +{
      MASTER001 => +{ ... },
      MASTER002 => +{ ... },
      MASTER003 => +{ ... },
    },
  });

  my @keys = ( 3 .. 8 );
  my %node_keys = $resolver->resolve_node_keys( 'MASTER', \@keys );
  ### %node_keys = ( MASTER001 => [ 3, 6 ], MASTER002 => [ 4, 7 ], MASTER003 => [ 5, 7 ] )
  while ( my ( $node, $keys ) = each %node_keys ) {
      process_node( $node, $keys );
  }

disconnect_all()

Disconnect all cached database handles.

cluster_info($cluster)

Return cluster info hash ref.

clusters($cluster)

Retrieve cluster member node names as Array.

  my $r = DBIx::DBHResolver->new;
  $r->config(+{ ... });
  my $cluster_or_node = 'activities_master';
  if ( $r->is_cluster($cluster_or_node) ) {
    for ($r->cluster( $cluster_or_node )) {
      process_activities_node($_);
    }
  }
  else {
    process_activities_node($cluster_or_node);
  }

is_cluster($cluster)

Return boolean value which cluster or not given name.

is_node($node)

Return boolean value which node or not given name.

GLOBAL VARIABLES ^

$CONFIG

Stored config on using class module.

$DBI

DBI module name, default 'DBI'. If you want to use custom DBI sub class, then you must override this variable.

$DBI_CONNECT_METHOD

DBI connect method name, default 'connect';

If you want to use DBIx::Connector instead of DBI, then:

  use DBIx::Connector;
  use DBIx::DBHResolver;

  $DBIx::DBHResolver::DBI = 'DBIx::Connector';
  $DBIx::DBHResolver::DBI_CONNECT_METHOD = 'new';
  $DBIx::DBHResolver::DBI_CONNECT_CACHED_METHOD = 'new';

  my $r = DBIx::DBHResolver->new;
  $r->config(+{...});

  $r->connect('main_master')->txn(
    fixup => sub {
      my $dbh = shift;
      ...
    }
  );

$DBI_CONNECT_CACHED_METHOD

DBI connect method name, default 'connect_cached';

AUTHOR ^

Kosuke Arisawa <arisawa@gmail.com>
Toru Yamaguchi <zigorou@cpan.org>

SEE ALSO ^

DBI

LICENSE ^

This library is free software; you can redistribute it and/or modify it under the same terms as Perl itself.

syntax highlighting: