<?xml version="1.0" encoding="utf-8" ?>
<!DOCTYPE chapter SYSTEM "chapter.dtd">
<chapter>
<header>
<copyright>
<year>1997</year><year>2013</year>
<holder>Ericsson AB. All Rights Reserved.</holder>
</copyright>
<legalnotice>
The contents of this file are subject to the Erlang Public License,
Version 1.1, (the "License"); you may not use this file except in
compliance with the License. You should have received a copy of the
Erlang Public License along with this software. If not, it can be
retrieved online at http://www.erlang.org/.
Software distributed under the License is distributed on an "AS IS"
basis, WITHOUT WARRANTY OF ANY KIND, either express or implied. See
the License for the specific language governing rights and limitations
under the License.
</legalnotice>
<title>Getting Started with Mnesia</title>
<prepared>Claes Wikström, Hans Nilsson and Håkan Mattsson</prepared>
<responsible></responsible>
<docno></docno>
<approved></approved>
<checked></checked>
<date></date>
<rev>C</rev>
<file>Mnesia_chap2.xml</file>
</header>
<p>This chapter introduces Mnesia. Following a brief discussion
about the first initial setup, a Mnesia database example is
demonstrated. This database example will be referenced in the
following chapters, where this example is modified in order to
illustrate various program constructs. In this chapter, the
following mandatory procedures are illustrated by examples:
</p>
<list type="bulleted">
<item>Starting an Erlang session and specifying a directory for the
Mnesia database.
</item>
<item>Initializing a database schema.
</item>
<item>Starting Mnesia and creating the required tables.</item>
</list>
<section>
<title>Starting Mnesia for the first time</title>
<p>Following is a simplified demonstration of a Mnesia system startup. This is the dialogue from the Erlang
shell:
</p>
<pre><![CDATA[
unix> erl -mnesia dir '"/tmp/funky"'
Erlang (BEAM) emulator version 4.9
Eshell V4.9 (abort with ^G)
1>
1> mnesia:create_schema([node()]).
ok
2> mnesia:start().
ok
3> mnesia:create_table(funky, []).
{atomic,ok}
4> mnesia:info().
---> Processes holding locks <---
---> Processes waiting for locks <---
---> Pending (remote) transactions <---
---> Active (local) transactions <---
---> Uncertain transactions <---
---> Active tables <---
funky : with 0 records occupying 269 words of mem
schema : with 2 records occupying 353 words of mem
===> System info in version "1.0", debug level = none <===
opt_disc. Directory "/tmp/funky" is used.
use fall-back at restart = false
running db nodes = [nonode@nohost]
stopped db nodes = []
remote = []
ram_copies = [funky]
disc_copies = [schema]
disc_only_copies = []
[{nonode@nohost,disc_copies}] = [schema]
[{nonode@nohost,ram_copies}] = [funky]
1 transactions committed, 0 aborted, 0 restarted, 1 logged to disc
0 held locks, 0 in queue; 0 local transactions, 0 remote
0 transactions waits for other nodes: []
ok
]]></pre>
<p>In the example above the following actions were performed:
</p>
<list type="bulleted">
<item>The Erlang system was started from the UNIX prompt
with a flag <c>-mnesia dir '"/tmp/funky"'</c>. This flag indicates
to Mnesia which directory will store the data.
</item>
<item>A new empty schema was initialized on the local node by evaluating
<c>mnesia:create_schema([node()]).</c> The schema contains
information about the database in general. This will be
thoroughly explained later on.
</item>
<item>The DBMS was started by evaluating <c>mnesia:start()</c>.
</item>
<item>A first table was created, called <c>funky</c> by evaluating
the expression <c>mnesia:create_table(funky, [])</c>. The table
was given default properties.
</item>
<item><c>mnesia:info()</c> was evaluated and subsequently displayed
information regarding the status of the database on the terminal.
</item>
</list>
</section>
<section>
<title>An Introductory Example</title>
<p>A Mnesia database is organized as a set of tables.
Each table is populated with instances (Erlang records).
A table also has a number of properties, such as location and
persistence.
</p>
<p>In this example we shall:
</p>
<list type="bulleted">
<item>Start an Erlang system, and specify the directory where
the database will be located.
</item>
<item>Initiate a new schema with an attribute that specifies
on which node, or nodes, the database will operate.
</item>
<item>Start Mnesia itself.
</item>
<item>Create and populate the database tables.
</item>
</list>
<section>
<title>The Example Database</title>
</section>
<p>In this database example, we will create the database and
relationships depicted in the following diagram. We will call this
database the <em>Company</em> database.
</p>
<image file="company.gif">
<icaption>Company Entity-Relation Diagram</icaption>
</image>
<p>The database model looks as follows:
</p>
<list type="bulleted">
<item>There are three entities: employee, project, and
department.
</item>
<item>
<p>There are three relationships between these entities:</p>
<list type="ordered">
<item>A department is managed by an employee, hence the
<em>manager</em> relationship.
</item>
<item>An employee works at a department, hence the
<em>at_dep</em> relationship.
</item>
<item>Each employee works on a number of projects, hence
the <em>in_proj</em> relationship.
</item>
</list>
</item>
</list>
<section>
<title>Defining Structure and Content</title>
<p>We first enter our record definitions into a text file
named <c>company.hrl</c>. This file defines the following
structure for our sample database:
</p>
<codeinclude file="company.hrl" tag="%0" type="erl"></codeinclude>
<p>The structure defines six tables in our database. In Mnesia,
the function <c>mnesia:create_table(Name, ArgList)</c> is
used to create tables. <c>Name</c> is the table
name <em>Note:</em> The current version of Mnesia does
not require that the name of the table is the same as the record
name, See Chapter 4:
<seealso marker="Mnesia_chap4#recordnames_tablenames">Record Names Versus Table Names.</seealso></p>
<p>For example, the table
for employees will be created with the function
<c>mnesia:create_table(employee, [{attributes, record_info(fields, employee)}]).</c> The table
name <c>employee</c> matches the name for records specified
in <c>ArgList</c>. The expression <c>record_info(fields, RecordName)</c> is processed by the Erlang preprocessor and
evaluates to a list containing the names of the different
fields for a record.
</p>
</section>
<section>
<title>The Program</title>
<p>The following shell interaction starts Mnesia and
initializes the schema for our <c>company</c> database:
</p>
<pre>
% <input>erl -mnesia dir '"/ldisc/scratch/Mnesia.Company"'</input>
Erlang (BEAM) emulator version 4.9
Eshell V4.9 (abort with ^G)
1> mnesia:create_schema([node()]).
ok
2> mnesia:start().
ok
</pre>
<p>The following program module creates and populates previously defined tables:
</p>
<codeinclude file="company.erl" tag="%0" type="erl"></codeinclude>
</section>
<section>
<title>The Program Explained</title>
<p>The following commands and functions were used to initiate the
Company database:
</p>
<list type="bulleted">
<item><c>% erl -mnesia dir '"/ldisc/scratch/Mnesia.Company"'.</c> This is a UNIX
command line entry which starts the Erlang system. The flag
<c>-mnesia dir Dir</c> specifies the location of the
database directory. The system responds and waits for
further input with the prompt <em>1></em>.
</item>
<item><c>mnesia:create_schema([node()]).</c> This function
has the format <c>mnesia:create_schema(DiscNodeList)</c> and
initiates a new schema. In this example, we have created a
non-distributed system using only one node. Schemas are fully
explained in Chapter 3:<seealso marker="Mnesia_chap3#def_schema">Defining a Schema</seealso>.
</item>
<item><c>mnesia:start().</c> This function starts
Mnesia. This function is fully explained in Chapter 3:
<seealso marker="Mnesia_chap3#start_mnesia">Starting Mnesia</seealso>.
</item>
</list>
<p>Continuing the dialogue with the Erlang shell will produce the following:</p>
<pre><![CDATA[
3> company:init().
{atomic,ok}
4> mnesia:info().
---> Processes holding locks <---
---> Processes waiting for locks <---
---> Pending (remote) transactions <---
---> Active (local) transactions <---
---> Uncertain transactions <---
---> Active tables <---
in_proj : with 0 records occuping 269 words of mem
at_dep : with 0 records occuping 269 words of mem
manager : with 0 records occuping 269 words of mem
project : with 0 records occuping 269 words of mem
dept : with 0 records occuping 269 words of mem
employee : with 0 records occuping 269 words of mem
schema : with 7 records occuping 571 words of mem
===> System info in version "1.0", debug level = none <===
opt_disc. Directory "/ldisc/scratch/Mnesia.Company" is used.
use fall-back at restart = false
running db nodes = [nonode@nohost]
stopped db nodes = []
remote = []
ram_copies =
[at_dep,dept,employee,in_proj,manager,project]
disc_copies = [schema]
disc_only_copies = []
[{nonode@nohost,disc_copies}] = [schema]
[{nonode@nohost,ram_copies}] =
[employee,dept,project,manager,at_dep,in_proj]
6 transactions committed, 0 aborted, 0 restarted, 6 logged to disc
0 held locks, 0 in queue; 0 local transactions, 0 remote
0 transactions waits for other nodes: []
ok
]]></pre>
<p>A set of tables is created:
</p>
<list type="bulleted">
<item><c>mnesia:create_table(Name,ArgList)</c>. This
function is used to create the required database tables. The
options available with <c>ArgList</c> are explained in
Chapter 3: <seealso marker="Mnesia_chap3#create_tables">Creating New Tables</seealso>. </item>
</list>
<p>The <c>company:init/0</c> function creates our tables. Two tables are
of type <c>bag</c>. This is the <c>manager</c> relation as well
the <c>in_proj</c> relation. This shall be interpreted as: An
employee can be manager over several departments, and an employee
can participate in several projects. However, the <c>at_dep</c>
relation is <c>set</c> because an employee can only work in one department.
In this data model we have examples of relations that are one-to-one (<c>set</c>),
as well as one-to-many (<c>bag</c>).
</p>
<p><c>mnesia:info()</c> now indicates that a database
which has seven local tables, of which, six are our
user defined tables and one is the schema.
Six transactions have been committed, as six successful transactions were run when
creating the tables.
</p>
<p>To write a function which inserts an employee record into the database, there must be an
<c>at_dep</c> record and a set of <c>in_proj</c> records inserted. Examine the following
code used to complete this action:
</p>
<codeinclude file="company.erl" tag="%1" type="erl"></codeinclude>
<list type="bulleted">
<item>
<p><c>insert_emp(Emp, DeptId, ProjNames) -></c>. The
<c>insert_emp/3</c> arguments are:</p>
<list type="ordered">
<item><c>Emp</c> is an employee record.
</item>
<item><c>DeptId</c> is the identity of the department where the employee is working.
</item>
<item><c>ProjNames</c> is a list of the names of the projects where the employee are working.</item>
</list>
</item>
</list>
<p>The <c>insert_emp(Emp, DeptId, ProjNames) -></c> function
creates a <em>functional object</em>. Functional objects
are identified by the term <c>Fun</c>. The Fun is passed
as a single argument to the function
<c>mnesia:transaction(Fun)</c>. This means that Fun is
run as a transaction with the following properties:
</p>
<list type="bulleted">
<item>Fun either succeeds or fails completely.
</item>
<item>Code which manipulates the same data records can be
run concurrently without the different processes interfering
with each other.
</item>
</list>
<p>The function can be used as:</p>
<code type="none">
Emp = #employee{emp_no= 104732,
name = klacke,
salary = 7,
sex = male,
phone = 98108,
room_no = {221, 015}},
insert_emp(Me, 'B/SFR', [Erlang, mnesia, otp]).
</code>
<note>
<p>Functional Objects (Funs) are described in the
Erlang Reference Manual, "Fun Expressions".
</p>
</note>
</section>
<section>
<title>Initial Database Content</title>
<p>After the insertion of the employee named <c>klacke</c>
we have the following records in the database:
</p>
<marker id="table2_1"></marker>
<table>
<row>
<cell align="left" valign="middle">emp_no</cell>
<cell align="left" valign="middle">name</cell>
<cell align="left" valign="middle">salary</cell>
<cell align="left" valign="middle">sex</cell>
<cell align="left" valign="middle">phone</cell>
<cell align="left" valign="middle">room_no</cell>
</row>
<row>
<cell align="left" valign="middle">104732</cell>
<cell align="left" valign="middle">klacke</cell>
<cell align="left" valign="middle">7</cell>
<cell align="left" valign="middle">male</cell>
<cell align="left" valign="middle">99586</cell>
<cell align="left" valign="middle">{221, 015}</cell>
</row>
<tcaption>
Employee</tcaption>
</table>
<p>An employee record has the following Erlang record/tuple
representation: <c>{employee, 104732, klacke, 7, male, 98108, {221, 015}}</c></p>
<marker id="table2_2"></marker>
<table>
<row>
<cell align="left" valign="middle">emp</cell>
<cell align="left" valign="middle">dept_name</cell>
</row>
<row>
<cell align="left" valign="middle">klacke</cell>
<cell align="left" valign="middle">B/SFR</cell>
</row>
<tcaption>
At_dep</tcaption>
</table>
<p>At_dep has the following Erlang tuple representation:
<c>{at_dep, klacke, 'B/SFR'}</c>.
</p>
<marker id="table3_3"></marker>
<table>
<row>
<cell align="left" valign="middle">emp</cell>
<cell align="left" valign="middle">proj_name</cell>
</row>
<row>
<cell align="left" valign="middle">klacke</cell>
<cell align="left" valign="middle">Erlang</cell>
</row>
<row>
<cell align="left" valign="middle">klacke</cell>
<cell align="left" valign="middle">otp</cell>
</row>
<row>
<cell align="left" valign="middle">klacke</cell>
<cell align="left" valign="middle">mnesia</cell>
</row>
<tcaption>
In_proj</tcaption>
</table>
<p>In_proj has the following Erlang tuple representation:
<c>{in_proj, klacke, 'Erlang', klacke, 'otp', klacke, 'mnesia'}</c></p>
<p>There is no difference between rows in a table and Mnesia
records. Both concepts are the same and will be used
interchangeably throughout this book.
</p>
<p>A Mnesia table is populated by Mnesia records. For example,
the tuple <c>{boss, klacke, bjarne}</c> is a record. The
second element in this tuple is the key. In order to uniquely
identify a table row both the key and the table name is
needed. The term <em>object identifier</em>,
(oid) is sometimes used for the arity two tuple {Tab, Key}. The oid for
the <c>{boss, klacke, bjarne}</c> record is the arity two
tuple <c>{boss, klacke}</c>. The first element of the tuple is
the type of the record and the second element is the key. An
oid can lead to zero, one, or more records depending on
whether the table type is <c>set</c> or <c>bag</c>.
</p>
<p>We were also able to insert the <c>{boss, klacke, bjarne}</c> record which contains an implicit reference to
another employee which does not yet exist in the
database. Mnesia does not enforce this.
</p>
</section>
<section>
<title>Adding Records and Relationships to the Database</title>
<p>After adding additional record to the Company database, we
may end up with the following records:
</p>
<p><em>Employees</em></p>
<code type="none">
{employee, 104465, "Johnson Torbjorn", 1, male, 99184, {242,038}}.
{employee, 107912, "Carlsson Tuula", 2, female,94556, {242,056}}.
{employee, 114872, "Dacker Bjarne", 3, male, 99415, {221,035}}.
{employee, 104531, "Nilsson Hans", 3, male, 99495, {222,026}}.
{employee, 104659, "Tornkvist Torbjorn", 2, male, 99514, {222,022}}.
{employee, 104732, "Wikstrom Claes", 2, male, 99586, {221,015}}.
{employee, 117716, "Fedoriw Anna", 1, female,99143, {221,031}}.
{employee, 115018, "Mattsson Hakan", 3, male, 99251, {203,348}}.
</code>
<p><em>Dept</em></p>
<code type="none">
{dept, 'B/SF', "Open Telecom Platform"}.
{dept, 'B/SFP', "OTP - Product Development"}.
{dept, 'B/SFR', "Computer Science Laboratory"}.
</code>
<p><em>Projects</em></p>
<code type="none">
%% projects
{project, erlang, 1}.
{project, otp, 2}.
{project, beam, 3}.
{project, mnesia, 5}.
{project, wolf, 6}.
{project, documentation, 7}.
{project, www, 8}.
</code>
<p>The above three tables, titled <c>employees</c>,
<c>dept</c>, and <c>projects</c>, are the tables which are
made up of real records. The following database content is
stored in the tables which is built on
relationships. These tables are titled <c>manager</c>,
<c>at_dep</c>, and <c>in_proj</c>.
</p>
<p><em>Manager</em></p>
<code type="none">
{manager, 104465, 'B/SF'}.
{manager, 104465, 'B/SFP'}.
{manager, 114872, 'B/SFR'}.
</code>
<p><em>At_dep</em></p>
<code type="none">
{at_dep, 104465, 'B/SF'}.
{at_dep, 107912, 'B/SF'}.
{at_dep, 114872, 'B/SFR'}.
{at_dep, 104531, 'B/SFR'}.
{at_dep, 104659, 'B/SFR'}.
{at_dep, 104732, 'B/SFR'}.
{at_dep, 117716, 'B/SFP'}.
{at_dep, 115018, 'B/SFP'}.
</code>
<p><em>In_proj</em></p>
<code type="none">
{in_proj, 104465, otp}.
{in_proj, 107912, otp}.
{in_proj, 114872, otp}.
{in_proj, 104531, otp}.
{in_proj, 104531, mnesia}.
{in_proj, 104545, wolf}.
{in_proj, 104659, otp}.
{in_proj, 104659, wolf}.
{in_proj, 104732, otp}.
{in_proj, 104732, mnesia}.
{in_proj, 104732, erlang}.
{in_proj, 117716, otp}.
{in_proj, 117716, documentation}.
{in_proj, 115018, otp}.
{in_proj, 115018, mnesia}.
</code>
<p>The room number is an attribute of the employee
record. This is a structured attribute which consists of a
tuple. The first element of the tuple identifies a corridor,
and the second element identifies the actual room in the
corridor. We could have chosen to represent this as a record
<c>-record(room, {corr, no}).</c> instead of an anonymous
tuple representation.
</p>
<p>The Company database is now initialized and contains
data. </p>
</section>
<section>
<title>Writing Queries</title>
<p>Retrieving data from DBMS should usually be done with <c>mnesia:read/3</c> or
<c>mnesia:read/1</c> functions. The following function raises the salary:</p>
<codeinclude file="company.erl" tag="%5" type="erl"></codeinclude>
<p>Since we want to update the record using <c>mnesia:write/1</c> after we have
increased the salary we acquire a write lock (third argument to read) when we read the
record from the table.
</p>
<p>It is not always the case that we can directly read the values from the table,
we might need to search the table or several tables to get the data we want, this
is done by writing database queries. Queries are always more expensive operations
than direct lookups done with <c>mnesia:read</c> and should be avoided in performance
critical code.</p>
<p>There are two methods for writing database queries:
</p>
<list type="bulleted">
<item>Mnesia functions
</item>
<item>QLC</item>
</list>
<section>
<title>Mnesia functions </title>
<p></p>
<p>The following function extracts the names of the female employees
stored in the database:
</p>
<pre>
mnesia:select(employee, [{#employee{sex = female, name = '$1', _ = '_'},[], ['$1']}]).
</pre>
<p>Select must always run within an activity such as a
transaction. To be able to call from the shell we might
construct a function as:
</p>
<codeinclude file="company.erl" tag="%20" type="erl"></codeinclude>
<p>The select expression matches all entries in table employee with
the field sex set to female.
</p>
<p>This function can be called from the shell as follows:
</p>
<pre>
(klacke@gin)1> <input>company:all_females().</input>
{atomic, ["Carlsson Tuula", "Fedoriw Anna"]}
</pre>
<p>See also the <seealso marker="Mnesia_chap4#matching">Pattern Matching </seealso>
chapter for a description of select and its syntax.
</p>
</section>
<section>
<title>Using QLC </title>
<p>This section contains simple introductory examples
only. Refer to <em>QLC reference manual</em> for a
full description of the QLC query language. Using QLC
might be more expensive than using Mnesia functions directly but
offers a nice syntax.
</p>
<p>The following function extracts a list of female employees
from the database:
</p>
<pre>
Q = qlc:q([E#employee.name || E <![CDATA[<-]]> mnesia:table(employee),
E#employee.sex == female]),
qlc:e(Q),
</pre>
<p>Accessing mnesia tables from a QLC list comprehension must
always be done within a transaction. Consider the following
function:
</p>
<codeinclude file="company.erl" tag="%2" type="erl"></codeinclude>
<p>This function can be called from the shell as follows:
</p>
<pre>
(klacke@gin)1> <input>company:females().</input>
{atomic, ["Carlsson Tuula", "Fedoriw Anna"]}
</pre>
<p>In traditional relational database terminology, the above
operation would be called a selection, followed by a projection.
</p>
<p>The list comprehension expression shown above contains a
number of syntactical elements.
</p>
<list type="bulleted">
<item>the first <c>[</c> bracket should be read as "build the
list"
</item>
<item>the <c>||</c> "such that" and the arrow <c><![CDATA[<-]]></c> should
be read as "taken from"
</item>
</list>
<p>Hence, the above list comprehension demonstrates the
formation of the list <c>E#employee.name</c> such that <c>E</c> is
taken from the table of employees and the <c>sex</c> attribute
of each records is equal with the atom <c>female</c>.
</p>
<p>The whole list comprehension must be given to the
<c>qlc:q/1</c> function.
</p>
<p>It is possible to combine list comprehensions with low
level Mnesia functions in the same transaction. If we want to
raise the salary of all female employees we execute:
</p>
<codeinclude file="company.erl" tag="%4" type="erl"></codeinclude>
<p>The function <c>raise_females/1</c> returns the tuple
<c>{atomic, Number}</c>, where <c>Number</c> is the number of
female employees who received a salary increase. Should an error
occur, the value <c>{aborted, Reason}</c> is returned. In the
case of an error, Mnesia guarantees that the salary is not
raised for any employees at all.
</p>
<pre>
33><input>company:raise_females(33).</input>
{atomic,2}
</pre>
</section>
</section>
</section>
</chapter>
