1DBD::Multi(3) User Contributed Perl Documentation DBD::Multi(3)
2
6 DBD::Multi - Manage Multiple Data Sources with Failover and Load
7 Balancing
8
10 use DBI;
11 my $other_dbh = DBI->connect(...);
13 my $dbh = DBI->connect( 'dbi:Multi:', undef, undef, {
15 dsns => [ # in priority order
16 10 => [ 'dbi:SQLite:read_one.db', '', '' ],
17 10 => [ 'dbi:SQLite:read_two.db', '', '' ],
18 20 => [ 'dbi:SQLite:master.db', '', '' ],
19 30 => $other_dbh,
20 40 => sub { DBI->connect },
21 ],
22 # optional
23 failed_max => 1, # short credibility
24 failed_expire => 60*60, # long memory
25 timeout => 10, # time out connection attempts after 10 seconds.
26 });
27 $dbh->prepare(...); # Works like any other DBI handle.
29 $dbh->multi_do_all( # Loops through every single DB handle.
32 sub {
33 my $dbh = shift;
34 ...
35 } );
36
38 This software manages multiple database connections for failovers and
39 also simple load balancing. It acts as a proxy between your code and
40 your database connections, transparently choosing a connection for each
41 query, based on your preferences and present availability of the DB
42 server.
43 This module is intended for read-only operations (where some other
45 application is being used to handle replication).
46 This software does not prevent write operations from being executed.
48 This is left up to the user. See "SUGGESTED USES" below for ideas.
49 The interface is nearly the same as other DBI drivers except that it
51 allows you to specify multiple connections for a single handle.
52
54 Configuring DSNs
55 Specify an attribute to the "connect()" constructor, "dsns". This is a
56 list of DSNs to configure. The configuration is given in pairs. First
57 comes the priority of the DSN. Second is the DSN.
58 The priorities specify which connections should be used first (lowest
60 to highest). As long as the lowest priority connection is responding,
61 the higher priority connections will never be used. If multiple
62 connections have the same priority, then one connection will be chosen
63 randomly for each operation. Note that the random DB is chosen when
64 the statement is prepared. Therefore executing multiple queries on
65 the same prepared statement handle will always run on the same
66 connection.
67 The second parameter can a DBI object, a code ref which returns a DBI
69 object, or a list of parameters to pass to the DBI "connect()"
70 instructor. If a set of parameters or a code ref is given, then
71 DBD::Multi will be able to attempt re-connect in the event that the
72 connection is lost. If a DBI object is used, the DBD::Multi will give
73 up permanently once that connection is lost.
74 These connections are lazy loaded, meaning they aren't made until they
76 are actually used.
77 Configuring Failures
79 By default, after a data source fails three times, it will not be tried
80 again for 5 minutes. After that period, the data source will be tried
81 again for future requests until it reaches its three failure limit (the
82 cycle repeats forever).
83 To change the maximum number of failures allowed before a data source
85 is deemed failed, set the "failed_max" parameter. To change the amount
86 of time we remember a data source as being failed, set the
87 "failed_expire" parameter in seconds.
88 Timing out connections.
90 By default, if you attempt to connect to an IP that isn't answering,
91 DBI will hang for a very long period of time. This behavior is not
92 desirable in a multi database setup. Instead, it is better to give up
93 on slow connections and move on to other databases quickly.
94 DBD::Multi will give up on connection attempts after 5 seconds and then
96 try another connection. You may set the "timeout" parameter to change
97 the timeout time, or set it to 0 to disable the timeout feature
98 completely.
99
101 multi_do_all
102 Loops through every database handle, executing an arbitrary coderef
103 passing the current database handle as the first parameter and the
104 original connection parameters as the second parameter.
105 If a database is unreachable, multi_do_all will skip over it.
107 use Data::Dumper;
109 my $expected_value = ...;
110 $dbh->multi_do_all(
111 sub {
112 my $dbh = shift;
113 my $source = shift;
114 my($value) = $dbh->selectrow_array("SELECT ...");
115 unless ( $value eq $expected_value ) {
116 die "Unexpected value, $value found. (Expected $expected_value). Data Source:\n", Dumper( $source );
117 }
118 } );
119
121 Here are some ideas on how to use this module effectively and safely.
122 It is important to remember that "DBD::Multi" is not intended for read-
124 write operations. One suggestion to prevent accidental write
125 operations is to make sure that the user you are connecting to the
126 databases with has privileges sufficiently restricted to prevent
127 updates.
128 Read-write operations should happen through a separate database handle
130 that will somehow trigger replication to all of your databases. For
131 example, your read-write handle might be connected to the master server
132 that replicates itself to all of the subordinate servers.
133 Read-only database calls within your application would be updated to
135 explicitly use the read-only (DBD::Multi) handle. It is not necessary
136 to find every single call that can be load balanced, since they can
137 safely be sent through the read/write handle as well.
138
140 There really isn't much of a TODO list for this module at this time.
141 Feel free to submit a bug report to github
142 <https://github.com/dwright/DBD-Multi/issues> if you think there is a
143 feature missing.
144 Although there is some code intended for read/write operations, this
146 should be considered not supported and not actively developed at this
147 time. The actual read/write code remains un-documented because in the
148 event that I ever do decide to work on supporting read/write
149 operations, the API is not guaranteed to stay the same. The focus of
150 this module is presently limited to read-only operations.
151
153 DBD::Multi has it's own suite of regression tests. But, suppose you
154 want to verify that you can slip DBD::Multi into whatever application
155 you already have written without breaking anything.
156 Thanks to a feature of DBI, you can regression test DBD::Multi using
158 any existing tests that already use DBI without having to update any of
159 your code. Simply set the environment variable DBI_AUTOPROXY to
160 'dbi:Multi:' and then run your tests. DBD::Multi should act as a
161 silent pipe between your application and whatever database driver you
162 were previously using. This will help you verify that you aren't
163 currently using some feature of the DBI that breaks DBD::Multi (If you
164 are, please do me a favor and submit a bug report so I can fix it).
165
167 There are other modules that have similar, but different objectives.
168 Depending on your specific needs these may be more or less suitable for
169 your task:
170 CGI::Application::Plugin::DBH
172 A plugin for the CGI::Application framework which makes it easy to
173 support two database handles, and also supports lazy-loading.
174 DBD::Multiplex
176 The original inspiration for this module. It doesn't support as
177 many connection configurations options at this module. It does
178 try to support write options in a single master, mutliple slave
179 configuration. It does this by parsing your SQL and trying to
180 decide if you were doing a read or write operation.
181 DBIx::HA
183 Written after this module. Built for high availability rather than
184 load balancing. It purposely ignores some DBI features in favor
185 of producing the fastest results for the most common operations.
186 It doesn't utilize the standard DBI->connect() API, which means it
187 will not work as a drop-in auto proxy.
188 DBI, perl - You should probably already know about these before using
190 this module.
191
193 Initially written by Casey West and Dan Wright for pair Networks, Inc.
194 (www.pair.com)
195 Maintained by Dan Wright. <DWRIGHT@CPAN.ORG>.
197perl v5.34.0 2021-07-22 DBD::Multi(3)