NAME

DBIx::Class::Storage::DBI::Replicated - BETA Replicated database support

SYNOPSIS

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', {balancer=>'::Random'}] );

## Add some slaves.  Basically this is an array of arrayrefs, where each
## arrayref is database connect information

$schema->storage->connect_replicants(
  [$dsn1, $user, $pass, \%opts],
  [$dsn2, $user, $pass, \%opts],
  [$dsn3, $user, $pass, \%opts],
);

## Now, just use the $schema as normal
$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<DBIx::Class::Storage::Replicated::Pool> and the replicants attribute for
## More. Also see transactions and L</execute_reliably> for alternative ways
## to force read traffic to the master.

DESCRIPTION

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 (INSERT, UPDATE, DELETE and even LAST_INSERT_ID) are routed to master database, all read-type queries (SELECTs) go to the slave database.

Basically, any method request that DBIx::Class::Storage::DBI would normally handle gets delegated to one of the two attributes: "read_handler" or to "write_handler". Additionally, some methods need to be distributed to all existing storages. This way our storage class is a drop in replacement for DBIx::Class::Storage::DBI.

Read traffic is spread across the replicants (slaves) occuring to a user selected algorithm. The default algorithm is random weighted.

NOTES

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.

REQUIREMENTS

Replicated Storage has additional requirements not currently part of DBIx::Class

Moose => 1.54
MooseX::AttributeHelpers => 0.12 
Moose::Util::TypeConstraints => 0.54
Class::MOP => 0.63

You will need to install these modules manually via CPAN or make them part of the Makefile for your distribution.

ATTRIBUTES

This class defines the following attributes.

schema

The underlying DBIx::Class::Schema object this storage is attaching

pool_type

Contains the classname which will instantiate the "pool" object. Defaults to: DBIx::Class::Storage::DBI::Replicated::Pool.

pool_args

Contains a hashref of initialized information to pass to the Balancer object. See DBIx::Class::Storage::Replicated::Pool for available arguments.

balancer_type

The replication pool requires a balance class to provider the methods for choose how to spread the query load across each replicant in the pool.

balancer_args

Contains a hashref of initialized information to pass to the Balancer object. See DBIx::Class::Storage::Replicated::Balancer for available arguments.

pool

Is a <DBIx::Class::Storage::DBI::Replicated::Pool> or derived class. This is a container class for one or more replicated databases.

balancer

Is a <DBIx::Class::Storage::DBI::Replicated::Balancer> or derived class. This is a class that takes a pool (<DBIx::Class::Storage::DBI::Replicated::Pool>)

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.

ATTRIBUTES IMPLEMENTING THE DBIx::Storage::DBI INTERFACE

The following methods are delegated all the methods required for the DBIx::Class::Storage::DBI interface.

read_handler

Defines an object that implements the read side of BIx::Class::Storage::DBI.

write_handler

Defines an object that implements the write side of BIx::Class::Storage::DBI.

METHODS

This class defines the following methods.

BUILDARGS

DBIx::Class::Schema 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.

_build_master

Lazy builder for the "master" attribute.

_build_pool

Lazy builder for the "pool" attribute.

_build_balancer

Lazy builder for the "balancer" attribute. This takes a Pool object so that the balancer knows which pool it's balancing.

_build_write_handler

Lazy builder for the "write_handler" attribute. The default is to set this to the "master".

_build_read_handler

Lazy builder for the "read_handler" attribute. The default is to set this to the "balancer".

around: connect_replicants

All calls to connect_replicants needs to have an existing $schema tacked onto top of the args, since DBIx::Storage::DBI needs it.

all_storages

Returns an array of of all the connected storage backends. The first element in the returned array is the master, and the remainings are each of the replicants.

execute_reliably ($coderef, ?@args)

Given a coderef, saves the current state of the "read_handler", 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.

set_reliable_storage

Sets the current $schema to be 'reliable', that is all queries, both read and write are sent to the master

set_balanced_storage

Sets the current $schema to be use the </balancer> for all reads, while all writea are sent to the master only

around: txn_do ($coderef)

Overload to the txn_do method, which is delegated to whatever the write_handler is set to. We overload this in order to wrap in inside a "execute_reliably" method.

connected

Check that the master and at least one of the replicants is connected.

ensure_connected

Make sure all the storages are connected.

limit_dialect

Set the limit_dialect for all existing storages

quote_char

Set the quote_char for all existing storages

name_sep

Set the name_sep for all existing storages

set_schema

Set the schema object for all existing storages

debug

set a debug flag across all storages

debugobj

set a debug object across all storages

debugfh

set a debugfh object across all storages

debugcb

set a debug callback across all storages

disconnect

disconnect everything

GOTCHAS

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.

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.

Otherwise, you can force a single query to use the master with the 'force_pool' attribute:

my $row = $resultset->search(undef, {force_pool=>'master'})->find($pk);

This attribute will safely be ignore by non replicated storages, so you can use the same code for both types of systems.

Lastly, you can use the "execute_reliably" method, which works very much like a transaction.

For debugging, you can turn replication on/off with the methods "set_reliable_storage" and "set_balanced_storage", 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.

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.

AUTHOR

John Napiorkowski <john.napiorkowski@takkle.com>

Based on code originated by:

Norbert Csongrádi <bert@cpan.org>
Peter Siklósi <einon@einon.hu>

LICENSE

You may distribute this code under the same terms as Perl itself.

cpanm DBIx::Class

perl -MCPAN -e shell
install DBIx::Class