<!DOCTYPE html PUBLIC "-//W3C//DTD HTML 4.01 Transitional//EN"> <html xmlns:fn="http://www.w3.org/2005/02/xpath-functions"> <head> <meta http-equiv="Content-Type" content="text/html; charset=UTF-8"> <link rel="stylesheet" href="../../../../doc/otp_doc.css" type="text/css"> <title>Erlang -- Transactions and Other Access Contexts</title> </head> <body bgcolor="white" text="#000000" link="#0000ff" vlink="#ff00ff" alink="#ff0000"><div id="container"> <script id="js" type="text/javascript" language="JavaScript" src="../../../../doc/js/flipmenu/flipmenu.js"></script><script id="js2" type="text/javascript" src="../../../../doc/js/erlresolvelinks.js"></script><script language="JavaScript" type="text/javascript"> <!-- function getWinHeight() { var myHeight = 0; if( typeof( window.innerHeight ) == 'number' ) { //Non-IE myHeight = window.innerHeight; } else if( document.documentElement && ( document.documentElement.clientWidth || document.documentElement.clientHeight ) ) { //IE 6+ in 'standards compliant mode' myHeight = document.documentElement.clientHeight; } else if( document.body && ( document.body.clientWidth || document.body.clientHeight ) ) { //IE 4 compatible myHeight = document.body.clientHeight; } return myHeight; } function setscrollpos() { var objf=document.getElementById('loadscrollpos'); document.getElementById("leftnav").scrollTop = objf.offsetTop - getWinHeight()/2; } function addEvent(obj, evType, fn){ if (obj.addEventListener){ obj.addEventListener(evType, fn, true); return true; } else if (obj.attachEvent){ var r = obj.attachEvent("on"+evType, fn); return r; } else { return false; } } addEvent(window, 'load', setscrollpos); //--></script><div id="leftnav"><div class="innertube"> <img alt="Erlang logo" src="../../../../doc/erlang-logo.png"><br><small><a href="users_guide.html">User's Guide</a><br><a href="index.html">Reference Manual</a><br><a href="release_notes.html">Release Notes</a><br><a href="../pdf/mnesia-4.4.13.pdf">PDF</a><br><a href="../../../../doc/index.html">Top</a></small><p><strong>Mnesia</strong><br><strong>User's Guide</strong><br><small>Version 4.4.13</small></p> <br><a href="javascript:openAllFlips()">Expand All</a><br><a href="javascript:closeAllFlips()">Contract All</a><p><small><strong>Chapters</strong></small></p> <ul class="flipMenu" imagepath="../../../../doc/js/flipmenu"> <li id="no" title="Introduction" expanded="false">Introduction<ul> <li><a href="Mnesia_chap1.html"> Top of chapter </a></li> <li title="About Mnesia"><a href="Mnesia_chap1.html#id2264416">About Mnesia</a></li> <li title="The Mnesia DataBase Management System (DBMS)"><a href="Mnesia_chap1.html#id2259328">The Mnesia DataBase Management System (DBMS)</a></li> </ul> </li> <li id="no" title="Getting Started with Mnesia" expanded="false">Getting Started with Mnesia<ul> <li><a href="Mnesia_chap2.html"> Top of chapter </a></li> <li title="Starting Mnesia for the first time"><a href="Mnesia_chap2.html#id2259207">Starting Mnesia for the first time</a></li> <li title="An Introductory Example"><a href="Mnesia_chap2.html#id2259513">An Introductory Example</a></li> </ul> </li> <li id="no" title="Building A Mnesia Database" expanded="false">Building A Mnesia Database<ul> <li><a href="Mnesia_chap3.html"> Top of chapter </a></li> <li title="Defining a Schema"><a href="Mnesia_chap3.html#id2268730">Defining a Schema</a></li> <li title="The Data Model"><a href="Mnesia_chap3.html#id2269025">The Data Model</a></li> <li title="Starting Mnesia"><a href="Mnesia_chap3.html#id2269927">Starting Mnesia</a></li> <li title="Creating New Tables"><a href="Mnesia_chap3.html#id2270354">Creating New Tables</a></li> </ul> </li> <li id="loadscrollpos" title="Transactions and Other Access Contexts" expanded="true">Transactions and Other Access Contexts<ul> <li><a href="Mnesia_chap4.html"> Top of chapter </a></li> <li title="Transaction Properties"><a href="Mnesia_chap4.html#id2270987">Transaction Properties</a></li> <li title="Locking"><a href="Mnesia_chap4.html#id2271248">Locking</a></li> <li title="Dirty Operations"><a href="Mnesia_chap4.html#id2271791">Dirty Operations</a></li> <li title="Record Names versus Table Names"><a href="Mnesia_chap4.html#id2272219">Record Names versus Table Names</a></li> <li title="Activity Concept and Various Access Contexts"><a href="Mnesia_chap4.html#id2272338">Activity Concept and Various Access Contexts</a></li> <li title="Nested transactions"><a href="Mnesia_chap4.html#id2272678">Nested transactions</a></li> <li title="Pattern Matching"><a href="Mnesia_chap4.html#id2272772">Pattern Matching</a></li> <li title="Iteration"><a href="Mnesia_chap4.html#id2273172">Iteration</a></li> </ul> </li> <li id="no" title="Miscellaneous Mnesia Features" expanded="false">Miscellaneous Mnesia Features<ul> <li><a href="Mnesia_chap5.html"> Top of chapter </a></li> <li title="Indexing"><a href="Mnesia_chap5.html#id2273564">Indexing</a></li> <li title="Distribution and Fault Tolerance"><a href="Mnesia_chap5.html#id2273702">Distribution and Fault Tolerance</a></li> <li title="Table Fragmentation"><a href="Mnesia_chap5.html#id2273885">Table Fragmentation</a></li> <li title="Local Content Tables"><a href="Mnesia_chap5.html#id2274960">Local Content Tables</a></li> <li title="Disc-less Nodes"><a href="Mnesia_chap5.html#id2274992">Disc-less Nodes</a></li> <li title="More Schema Management"><a href="Mnesia_chap5.html#id2275179">More Schema Management</a></li> <li title="Mnesia Event Handling"><a href="Mnesia_chap5.html#id2275329">Mnesia Event Handling</a></li> <li title="Debugging Mnesia Applications"><a href="Mnesia_chap5.html#id2275937">Debugging Mnesia Applications</a></li> <li title="Concurrent Processes in Mnesia"><a href="Mnesia_chap5.html#id2276097">Concurrent Processes in Mnesia</a></li> <li title="Prototyping"><a href="Mnesia_chap5.html#id2276148">Prototyping</a></li> <li title="Object Based Programming with Mnesia"><a href="Mnesia_chap5.html#id2276292">Object Based Programming with Mnesia</a></li> </ul> </li> <li id="no" title="Mnesia System Information" expanded="false">Mnesia System Information<ul> <li><a href="Mnesia_chap7.html"> Top of chapter </a></li> <li title="Database Configuration Data"><a href="Mnesia_chap7.html#id2276572">Database Configuration Data</a></li> <li title="Core Dumps"><a href="Mnesia_chap7.html#id2276614">Core Dumps</a></li> <li title="Dumping Tables"><a href="Mnesia_chap7.html#id2276638">Dumping Tables</a></li> <li title="Checkpoints"><a href="Mnesia_chap7.html#id2276678">Checkpoints</a></li> <li title="Files"><a href="Mnesia_chap7.html#id2276957">Files</a></li> <li title="Loading of Tables at Start-up"><a href="Mnesia_chap7.html#id2277368">Loading of Tables at Start-up</a></li> <li title="Recovery from Communication Failure"><a href="Mnesia_chap7.html#id2277560">Recovery from Communication Failure</a></li> <li title="Recovery of Transactions"><a href="Mnesia_chap7.html#id2277693">Recovery of Transactions</a></li> <li title="Backup, Fallback, and Disaster Recovery"><a href="Mnesia_chap7.html#id2277864">Backup, Fallback, and Disaster Recovery</a></li> </ul> </li> <li id="no" title="Combining Mnesia with SNMP" expanded="false">Combining Mnesia with SNMP<ul> <li><a href="Mnesia_chap8.html"> Top of chapter </a></li> <li title="Combining Mnesia and SNMP "><a href="Mnesia_chap8.html#id2278809">Combining Mnesia and SNMP </a></li> </ul> </li> <li id="no" title="Appendix A: Mnesia Error Messages" expanded="false">Appendix A: Mnesia Error Messages<ul> <li><a href="Mnesia_App_A.html"> Top of chapter </a></li> <li title="Errors in Mnesia"><a href="Mnesia_App_A.html#id2278969">Errors in Mnesia</a></li> </ul> </li> <li id="no" title="Appendix B: The Backup Call Back Interface" expanded="false">Appendix B: The Backup Call Back Interface<ul> <li><a href="Mnesia_App_B.html"> Top of chapter </a></li> <li title="mnesia_backup callback behavior"><a href="Mnesia_App_B.html#id2279207">mnesia_backup callback behavior</a></li> </ul> </li> <li id="no" title="Appendix C: The Activity Access Call Back Interface" expanded="false">Appendix C: The Activity Access Call Back Interface<ul> <li><a href="Mnesia_App_C.html"> Top of chapter </a></li> <li title="mnesia_access callback behavior"><a href="Mnesia_App_C.html#id2279378">mnesia_access callback behavior</a></li> </ul> </li> <li id="no" title="Appendix D: The Fragmented Table Hashing Call Back Interface" expanded="false">Appendix D: The Fragmented Table Hashing Call Back Interface<ul> <li><a href="Mnesia_App_D.html"> Top of chapter </a></li> <li title="mnesia_frag_hash callback behavior"><a href="Mnesia_App_D.html#id2279587">mnesia_frag_hash callback behavior</a></li> </ul> </li> </ul> </div></div> <div id="content"> <div class="innertube"> <h1>4 Transactions and Other Access Contexts</h1> <p>This chapter describes the Mnesia transaction system and the transaction properties which make Mnesia a fault tolerant, distributed database management system. </p> <p>Also covered in this chapter are the locking functions, including table locks and sticky locks, as well as alternative functions which bypass the transaction system in favor of improved speed and reduced overheads. These functions are called "dirty operations". We also describe the usage of nested transactions. This chapter contains the following sections: </p> <ul> <li>transaction properties, which include atomicity, consistency, isolation, and durability </li> <li>Locking </li> <li>Dirty operations </li> <li>Record names vs table names </li> <li>Activity concept and various access contexts </li> <li>Nested transactions </li> <li>Pattern matching </li> <li>Iteration </li> </ul> <h3><a name="id2270987">4.1 Transaction Properties</a></h3> <a name="trans_prop"></a> <p>Transactions are an important tool when designing fault tolerant, distributed systems. A Mnesia transaction is a mechanism by which a series of database operations can be executed as one functional block. The functional block which is run as a transaction is called a Functional Object (Fun), and this code can read, write, or delete Mnesia records. The Fun is evaluated as a transaction which either commits, or aborts. If a transaction succeeds in executing Fun it will replicate the action on all nodes involved, or abort if an error occurs. </p> <p>The following example shows a transaction which raises the salary of certain employee numbers. </p> <div class="example"><pre> raise(Eno, Raise) -> F = fun() -> [E] = mnesia:read(employee, Eno, write), Salary = E#employee.salary + Raise, New = E#employee{salary = Salary}, mnesia:write(New) end, mnesia:transaction(F).</pre></div> <p>The transaction <span class="code">raise(Eno, Raise) - ></span> contains a Fun made up of four lines of code. This Fun is called by the statement <span class="code">mnesia:transaction(F)</span> and returns a value. </p> <p>The Mnesia transaction system facilitates the construction of reliable, distributed systems by providing the following important properties: </p> <ul> <li>The transaction handler ensures that a Fun which is placed inside a transaction does not interfere with operations embedded in other transactions when it executes a series of operations on tables. </li> <li>The transaction handler ensures that either all operations in the transaction are performed successfully on all nodes atomically, or the transaction fails without permanent effect on any of the nodes. </li> <li>The Mnesia transactions have four important properties, which we call <strong>A</strong>tomicity, <strong>C</strong>onsistency,<strong>I</strong>solation, and <strong>D</strong>urability, or ACID for short. These properties are described in the following sub-sections.</li> </ul> <h4>Atomicity</h4> <p><strong>Atomicity</strong> means that database changes which are executed by a transaction take effect on all nodes involved, or on none of the nodes. In other words, the transaction either succeeds entirely, or it fails entirely. </p> <p>Atomicity is particularly important when we want to atomically write more than one record in the same transaction. The <span class="code">raise/2</span> function, shown as an example above, writes one record only. The <span class="code">insert_emp/3</span> function, shown in the program listing in Chapter 2, writes the record <span class="code">employee</span> as well as employee relations such as <span class="code">at_dep</span> and <span class="code">in_proj</span> into the database. If we run this latter code inside a transaction, then the transaction handler ensures that the transaction either succeeds completely, or not at all. </p> <p>Mnesia is a distributed DBMS where data can be replicated on several nodes. In many such applications, it is important that a series of write operations are performed atomically inside a transaction. The atomicity property ensures that a transaction take effect on all nodes, or none at all. </p> <h4>Consistency</h4> <p><strong>Consistency</strong>. This transaction property ensures that a transaction always leaves the DBMS in a consistent state. For example, Mnesia ensures that inconsistencies will not occur if Erlang, Mnesia or the computer crashes while a write operation is in progress. </p> <h4>Isolation</h4> <p><strong>Isolation</strong>. This transaction property ensures that transactions which execute on different nodes in a network, and access and manipulate the same data records, will not interfere with each other. </p> <p>The isolation property makes it possible to concurrently execute the <span class="code">raise/2</span> function. A classical problem in concurrency control theory is the so called "lost update problem". </p> <p>The isolation property is extremely useful if the following circumstances occurs where an employee (with an employee number 123) and two processes, (P1 and P2), are concurrently trying to raise the salary for the employee. The initial value of the employees salary is, for example, 5. Process P1 then starts to execute, it reads the employee record and adds 2 to the salary. At this point in time, process P1 is for some reason preempted and process P2 has the opportunity to run. P2 reads the record, adds 3 to the salary, and finally writes a new employee record with the salary set to 8. Now, process P1 start to run again and writes its employee record with salary set to 7, thus effectively overwriting and undoing the work performed by process P2. The update performed by P2 is lost. </p> <p>A transaction system makes it possible to concurrently execute two or more processes which manipulate the same record. The programmer does not need to check that the updates are synchronous, this is overseen by the transaction handler. All programs accessing the database through the transaction system may be written as if they had sole access to the data. </p> <h4>Durability</h4> <p><strong>Durability</strong>. This transaction property ensures that changes made to the DBMS by a transaction are permanent. Once a transaction has been committed, all changes made to the database are durable - i.e. they are written safely to disc and will not be corrupted or disappear. </p> <div class="note"> <div class="label">Note</div> <div class="content"><p> <p>The durability feature described does not entirely apply to situations where Mnesia is configured as a "pure" primary memory database. </p> </p></div> </div> <h3><a name="id2271248">4.2 Locking</a></h3> <p>Different transaction managers employ different strategies to satisfy the isolation property. Mnesia uses the standard technique of two-phase locking. This means that locks are set on records before they are read or written. Mnesia uses five different kinds of locks. </p> <ul> <li> <strong>Read locks</strong>. A read lock is set on one replica of a record before it can be read. </li> <li> <strong>Write locks</strong>. Whenever a transaction writes to an record, write locks are first set on all replicas of that particular record. </li> <li> <strong>Read table locks</strong>. If a transaction traverses an entire table in search for a record which satisfy some particular property, it is most inefficient to set read locks on the records, one by one. It is also very memory consuming, since the read locks themselves may take up considerable space if the table is very large. For this reason, Mnesia can set a read lock on an entire table. </li> <li> <strong>Write table locks</strong>. If a transaction writes a large number of records to one table, it is possible to set a write lock on the entire table. </li> <li> <strong>Sticky locks</strong>. These are write locks that stay in place at a node after the transaction which initiated the lock has terminated. </li> </ul> <p>Mnesia employs a strategy whereby functions such as <span class="code">mnesia:read/1</span> acquire the necessary locks dynamically as the transactions execute. Mnesia automatically sets and releases the locks and the programmer does not have to code these operations. </p> <p>Deadlocks can occur when concurrent processes set and release locks on the same records. Mnesia employs a "wait-die" strategy to resolve these situations. If Mnesia suspects that a deadlock can occur when a transaction tries to set a lock, the transaction is forced to release all its locks and sleep for a while. The Fun in the transaction will be evaluated one more time. </p> <p>For this reason, it is important that the code inside the Fun given to <span class="code">mnesia:transaction/1</span> is pure. Some strange results can occur if, for example, messages are sent by the transaction Fun. The following example illustrates this situation: </p> <div class="example"><pre> bad_raise(Eno, Raise) -> F = fun() -> [E] = mnesia:read({employee, Eno}), Salary = E#employee.salary + Raise, New = E#employee{salary = Salary}, io:format("Trying to write ... ~n", []), mnesia:write(New) end, mnesia:transaction(F).</pre></div> <p>This transaction could write the text <span class="code">"Trying to write ... "</span> a thousand times to the terminal. Mnesia does guarantee, however, that each and every transaction will eventually run. As a result, Mnesia is not only deadlock free, but also livelock free. </p> <p>The Mnesia programmer cannot prioritize one particular transaction to execute before other transactions which are waiting to execute. As a result, the Mnesia DBMS transaction system is not suitable for hard real time applications. However, Mnesia contains other features that have real time properties. </p> <p>Mnesia dynamically sets and releases locks as transactions execute, therefore, it is very dangerous to execute code with transaction side-effects. In particular, a <span class="code">receive</span> statement inside a transaction can lead to a situation where the transaction hangs and never returns, which in turn can cause locks not to release. This situation could bring the whole system to a standstill since other transactions which execute in other processes, or on other nodes, are forced to wait for the defective transaction. </p> <p>If a transaction terminates abnormally, Mnesia will automatically release the locks held by the transaction. </p> <p>We have shown examples of a number of functions that can be used inside a transaction. The following list shows the <strong>simplest</strong> Mnesia functions that work with transactions. It is important to realize that these functions must be embedded in a transaction. If no enclosing transaction (or other enclosing Mnesia activity) exists, they will all fail. </p> <ul> <li> <span class="code">mnesia:transaction(Fun) -> {aborted, Reason} |{atomic, Value}</span>. This function executes one transaction with the functional object <span class="code">Fun</span> as the single parameter. </li> <li> <span class="code">mnesia:read({Tab, Key}) -> transaction abort | RecordList</span>. This function reads all records with <span class="code">Key</span> as key from table <span class="code">Tab</span>. This function has the same semantics regardless of the location of <span class="code">Table</span>. If the table is of type <span class="code">bag</span>, the <span class="code">read({Tab, Key})</span> can return an arbitrarily long list. If the table is of type <span class="code">set</span>, the list is either of length one, or <span class="code">[]</span>. </li> <li> <span class="code">mnesia:wread({Tab, Key}) -> transaction abort | RecordList</span>. This function behaves the same way as the previously listed <span class="code">read/1</span> function, except that it acquires a write lock instead of a read lock. If we execute a transaction which reads a record, modifies the record, and then writes the record, it is slightly more efficient to set the write lock immediately. In cases where we issue a <span class="code">mnesia:read/1</span>, followed by a <span class="code">mnesia:write/1</span>, the first read lock must be upgraded to a write lock when the write operation is executed. </li> <li> <span class="code">mnesia:write(Record) -> transaction abort | ok</span>. This function writes a record into the database. The <span class="code">Record</span> argument is an instance of a record. The function returns <span class="code">ok</span>, or aborts the transaction if an error should occur. </li> <li> <span class="code">mnesia:delete({Tab, Key}) -> transaction abort | ok</span>. This function deletes all records with the given key. </li> <li> <span class="code">mnesia:delete_object(Record) -> transaction abort | ok</span>. This function deletes records with object id <span class="code">Record</span>. This function is used when we want to delete only some records in a table of type <span class="code">bag</span>. </li> </ul> <h4>Sticky Locks</h4> <p>As previously stated, the locking strategy used by Mnesia is to lock one record when we read a record, and lock all replicas of a record when we write a record. However, there are applications which use Mnesia mainly for its fault-tolerant qualities, and these applications may be configured with one node doing all the heavy work, and a standby node which is ready to take over in case the main node fails. Such applications may benefit from using sticky locks instead of the normal locking scheme. </p> <p>A sticky lock is a lock which stays in place at a node after the transaction which first acquired the lock has terminated. To illustrate this, assume that we execute the following transaction: </p> <div class="example"><pre> F = fun() -> mnesia:write(#foo{a = kalle}) end, mnesia:transaction(F). </pre></div> <p>The <span class="code">foo</span> table is replicated on the two nodes <span class="code">N1</span> and <span class="code">N2</span>. <br> Normal locking requires: </p> <ul> <li>one network rpc (2 messages) to acquire the write lock </li> <li>three network messages to execute the two-phase commit protocol. </li> </ul> <p>If we use sticky locks, we must first change the code as follows: </p> <div class="example"><pre> F = fun() -> mnesia:s_write(#foo{a = kalle}) end, mnesia:transaction(F). </pre></div> <p>This code uses the <span class="code">s_write/1</span> function instead of the <span class="code">write/1</span> function. The <span class="code">s_write/1</span> function sets a sticky lock instead of a normal lock. If the table is not replicated, sticky locks have no special effect. If the table is replicated, and we set a sticky lock on node <span class="code">N1</span>, this lock will then stick to node <span class="code">N1</span>. The next time we try to set a sticky lock on the same record at node <span class="code">N1</span>, Mnesia will see that the lock is already set and will not do a network operation in order to acquire the lock. </p> <p>It is much more efficient to set a local lock than it is to set a networked lock, and for this reason sticky locks can benefit application that use a replicated table and perform most of the work on only one of the nodes. </p> <p>If a record is stuck at node <span class="code">N1</span> and we try to set a sticky lock for the record on node <span class="code">N2</span>, the record must be unstuck. This operation is expensive and will reduce performance. The unsticking is done automatically if we issue <span class="code">s_write/1</span> requests at <span class="code">N2</span>. </p> <h4>Table Locks</h4> <p>Mnesia supports read and write locks on whole tables as a complement to the normal locks on single records. As previously stated, Mnesia sets and releases locks automatically, and the programmer does not have to code these operations. However, transactions which read and write a large number of records in a specific table will execute more efficiently if we start the transaction by setting a table lock on this table. This will block other concurrent transactions from the table. The following two function are used to set explicit table locks for read and write operations: </p> <ul> <li> <span class="code">mnesia:read_lock_table(Tab)</span> Sets a read lock on the table <span class="code">Tab</span> </li> <li> <span class="code">mnesia:write_lock_table(Tab)</span> Sets a write lock on the table <span class="code">Tab</span> </li> </ul> <p>Alternate syntax for acquisition of table locks is as follows: </p> <div class="example"><pre> mnesia:lock({table, Tab}, read) mnesia:lock({table, Tab}, write) </pre></div> <p>The matching operations in Mnesia may either lock the entire table or just a single record (when the key is bound in the pattern). </p> <h4>Global Locks</h4> <p>Write locks are normally acquired on all nodes where a replica of the table resides (and is active). Read locks are acquired on one node (the local one if a local replica exists). </p> <p>The function <span class="code">mnesia:lock/2</span> is intended to support table locks (as mentioned previously) but also for situations when locks need to be acquired regardless of how tables have been replicated: </p> <div class="example"><pre> mnesia:lock({global, GlobalKey, Nodes}, LockKind) LockKind ::= read | write | ... </pre></div> <p>The lock is acquired on the LockItem on all Nodes in the nodes list.</p> <h3><a name="id2271791">4.3 Dirty Operations</a></h3> <p>In many applications, the overhead of processing a transaction may result in a loss of performance. Dirty operation are short cuts which bypass much of the processing and increase the speed of the transaction. </p> <p>Dirty operation are useful in many situations, for example in a datagram routing application where Mnesia stores the routing table, and it is time consuming to start a whole transaction every time a packet is received. For this reason, Mnesia has functions which manipulate tables without using transactions. This alternative to processing is known as a dirty operation. However, it is important to realize the trade-off in avoiding the overhead of transaction processing: </p> <ul> <li>The atomicity and the isolation properties of Mnesia are lost. </li> <li>The isolation property is compromised, because other Erlang processes, which use transaction to manipulate the data, do not get the benefit of isolation if we simultaneously use dirty operations to read and write records from the same table. </li> </ul> <p>The major advantage of dirty operations is that they execute much faster than equivalent operations that are processed as functional objects within a transaction. </p> <p>Dirty operations are written to disc if they are performed on a table of type <span class="code">disc_copies</span>, or type <span class="code">disc_only_copies</span>. Mnesia also ensures that all replicas of a table are updated if a dirty write operation is performed on a table. </p> <p>A dirty operation will ensure a certain level of consistency. For example, it is not possible for dirty operations to return garbled records. Hence, each individual read or write operation is performed in an atomic manner. </p> <p>All dirty functions execute a call to <span class="code">exit({aborted, Reason})</span> on failure. Even if the following functions are executed inside a transaction no locks will be acquired. The following functions are available: </p> <ul> <li> <span class="code">mnesia:dirty_read({Tab, Key})</span>. This function reads record(s) from Mnesia. </li> <li> <span class="code">mnesia:dirty_write(Record)</span>. This function writes the record <span class="code">Record</span> </li> <li> <span class="code">mnesia:dirty_delete({Tab, Key})</span>. This function deletes record(s) with the key <span class="code">Key</span>. </li> <li> <span class="code">mnesia:dirty_delete_object(Record)</span> This function is the dirty operation alternative to the function <span class="code">delete_object/1</span> </li> <li> <p><span class="code">mnesia:dirty_first(Tab)</span>. This function returns the "first" key in the table <span class="code">Tab</span>. </p> <p>Records in <span class="code">set</span> or <span class="code">bag</span> tables are not sorted. However, there is a record order which is not known to the user. This means that it is possible to traverse a table by means of this function in conjunction with the <span class="code">dirty_next/2</span> function. </p> <p>If there are no records at all in the table, this function will return the atom <span class="code">'$end_of_table'</span>. It is not recommended to use this atom as the key for any user records. </p> </li> <li> <span class="code">mnesia:dirty_next(Tab, Key)</span>. This function returns the "next" key in the table <span class="code">Tab</span>. This function makes it possible to traverse a table and perform some operation on all records in the table. When the end of the table is reached the special key <span class="code">'$end_of_table'</span> is returned. Otherwise, the function returns a key which can be used to read the actual record. <br> The behavior is undefined if any process perform a write operation on the table while we traverse the table with the <span class="code">dirty_next/2</span> function. This is because <span class="code">write</span> operations on a Mnesia table may lead to internal reorganizations of the table itself. This is an implementation detail, but remember the dirty functions are low level functions. </li> <li> <span class="code">mnesia:dirty_last(Tab)</span> This function works exactly as <span class="code">mnesia:dirty_first/1</span> but returns the last object in Erlang term order for the <span class="code">ordered_set</span> table type. For all other table types, <span class="code">mnesia:dirty_first/1</span> and <span class="code">mnesia:dirty_last/1</span> are synonyms. </li> <li> <span class="code">mnesia:dirty_prev(Tab, Key)</span> This function works exactly as <span class="code">mnesia:dirty_next/2</span> but returns the previous object in Erlang term order for the ordered_set table type. For all other table types, <span class="code">mnesia:dirty_next/2</span> and <span class="code">mnesia:dirty_prev/2</span> are synonyms. </li> <li> <p><span class="code">mnesia:dirty_slot(Tab, Slot)</span></p> <p>Returns the list of records that are associated with Slot in a table. It can be used to traverse a table in a manner similar to the <span class="code">dirty_next/2</span> function. A table has a number of slots that range from zero to some unknown upper bound. The function <span class="code">dirty_slot/2</span> returns the special atom <span class="code">'$end_of_table'</span> when the end of the table is reached. <br> The behavior of this function is undefined if the table is written on while being traversed. <span class="code">mnesia:read_lock_table(Tab)</span> may be used to ensure that no transaction protected writes are performed during the iteration. </p> </li> <li> <p><span class="code">mnesia:dirty_update_counter({Tab,Key}, Val)</span>. </p> <p>Counters are positive integers with a value greater than or equal to zero. Updating a counter will add the <span class="code">Val</span> and the counter where <span class="code">Val</span> is a positive or negative integer. <br> There exists no special counter records in Mnesia. However, records on the form of <span class="code">{TabName, Key, Integer}</span> can be used as counters, and can be persistent. </p> <p>It is not possible to have transaction protected updates of counter records. </p> <p>There are two significant differences when using this function instead of reading the record, performing the arithmetic, and writing the record: </p> <ul> <li>it is much more efficient </li> <li>the <span class="code">dirty_update_counter/2</span> function is performed as an atomic operation although it is not protected by a transaction. Accordingly, no table update is lost if two processes simultaneously execute the <span class="code">dirty_update_counter/2</span> function. </li> </ul> </li> <li> <span class="code">mnesia:dirty_match_object(Pat)</span>. This function is the dirty equivalent of <span class="code">mnesia:match_object/1</span>. </li> <li> <span class="code">mnesia:dirty_select(Tab, Pat)</span>. This function is the dirty equivalent of <span class="code">mnesia:select/2</span>. </li> <li> <span class="code">mnesia:dirty_index_match_object(Pat, Pos)</span>. This function is the dirty equivalent of <span class="code">mnesia:index_match_object/2</span>. </li> <li> <span class="code">mnesia:dirty_index_read(Tab, SecondaryKey, Pos)</span>. This function is the dirty equivalent of <span class="code">mnesia:index_read/3</span>. </li> <li> <span class="code">mnesia:dirty_all_keys(Tab)</span>. This function is the dirty equivalent of <span class="code">mnesia:all_keys/1</span>. </li> </ul> <h3><a name="id2272219">4.4 Record Names versus Table Names</a></h3> <a name="recordnames_tablenames"></a> <p>In Mnesia, all records in a table must have the same name. All the records must be instances of the same record type. The record name does however not necessarily be the same as the table name. Even though that it is the case in the most of the examples in this document. If a table is created without the <span class="code">record_name</span> property the code below will ensure all records in the tables have the same name as the table: </p> <div class="example"><pre> mnesia:create_table(subscriber, []) </pre></div> <p>However, if the table is is created with an explicit record name as argument, as shown below, it is possible to store subscriber records in both of the tables regardless of the table names: </p> <div class="example"><pre> TabDef = [{record_name, subscriber}], mnesia:create_table(my_subscriber, TabDef), mnesia:create_table(your_subscriber, TabDef). </pre></div> <p>In order to access such tables it is not possible to use the simplified access functions as described earlier in the document. For example, writing a subscriber record into a table requires a <span class="code">mnesia:write/3</span>function instead of the simplified functions <span class="code">mnesia:write/1</span> and <span class="code">mnesia:s_write/1</span>: </p> <div class="example"><pre> mnesia:write(subscriber, #subscriber{}, write) mnesia:write(my_subscriber, #subscriber{}, sticky_write) mnesia:write(your_subscriber, #subscriber{}, write) </pre></div> <p>The following simplified piece of code illustrates the relationship between the simplified access functions used in most examples and their more flexible counterparts: </p> <div class="example"><pre> mnesia:dirty_write(Record) -> Tab = element(1, Record), mnesia:dirty_write(Tab, Record). mnesia:dirty_delete({Tab, Key}) -> mnesia:dirty_delete(Tab, Key). mnesia:dirty_delete_object(Record) -> Tab = element(1, Record), mnesia:dirty_delete_object(Tab, Record) mnesia:dirty_update_counter({Tab, Key}, Incr) -> mnesia:dirty_update_counter(Tab, Key, Incr). mnesia:dirty_read({Tab, Key}) -> Tab = element(1, Record), mnesia:dirty_read(Tab, Key). mnesia:dirty_match_object(Pattern) -> Tab = element(1, Pattern), mnesia:dirty_match_object(Tab, Pattern). mnesia:dirty_index_match_object(Pattern, Attr) Tab = element(1, Pattern), mnesia:dirty_index_match_object(Tab, Pattern, Attr). mnesia:write(Record) -> Tab = element(1, Record), mnesia:write(Tab, Record, write). mnesia:s_write(Record) -> Tab = element(1, Record), mnesia:write(Tab, Record, sticky_write). mnesia:delete({Tab, Key}) -> mnesia:delete(Tab, Key, write). mnesia:s_delete({Tab, Key}) -> mnesia:delete(Tab, Key, sticky_write). mnesia:delete_object(Record) -> Tab = element(1, Record), mnesia:delete_object(Tab, Record, write). mnesia:s_delete_object(Record) -> Tab = element(1, Record), mnesia:delete_object(Tab, Record. sticky_write). mnesia:read({Tab, Key}) -> mnesia:read(Tab, Key, read). mnesia:wread({Tab, Key}) -> mnesia:read(Tab, Key, write). mnesia:match_object(Pattern) -> Tab = element(1, Pattern), mnesia:match_object(Tab, Pattern, read). mnesia:index_match_object(Pattern, Attr) -> Tab = element(1, Pattern), mnesia:index_match_object(Tab, Pattern, Attr, read). </pre></div> <h3><a name="id2272338">4.5 Activity Concept and Various Access Contexts</a></h3> <p>As previously described, a functional object (Fun) performing table access operations as listed below may be passed on as arguments to the function <span class="code">mnesia:transaction/1,2,3</span>: </p> <ul> <li> <p>mnesia:write/3 (write/1, s_write/1)</p> </li> <li> <p>mnesia:delete/3 (delete/1, s_delete/1)</p> </li> <li> <p>mnesia:delete_object/3 (delete_object/1, s_delete_object/1)</p> </li> <li> <p>mnesia:read/3 (read/1, wread/1)</p> </li> <li> <p>mnesia:match_object/2 (match_object/1)</p> </li> <li> <p>mnesia:select/3 (select/2)</p> </li> <li> <p>mnesia:foldl/3 (foldl/4, foldr/3, foldr/4)</p> </li> <li> <p>mnesia:all_keys/1</p> </li> <li> <p>mnesia:index_match_object/4 (index_match_object/2)</p> </li> <li> <p>mnesia:index_read/3</p> </li> <li> <p>mnesia:lock/2 (read_lock_table/1, write_lock_table/1)</p> </li> <li> <p>mnesia:table_info/2</p> </li> </ul> <p>These functions will be performed in a transaction context involving mechanisms like locking, logging, replication, checkpoints, subscriptions, commit protocols etc.However, the same function may also be evaluated in other activity contexts. <br> The following activity access contexts are currently supported: </p> <ul> <li> <p>transaction </p> </li> <li> <p>sync_transaction</p> </li> <li> <p>async_dirty</p> </li> <li> <p>sync_dirty</p> </li> <li> <p>ets</p> </li> </ul> <p>By passing the same "fun" as argument to the function <span class="code">mnesia:sync_transaction(Fun [, Args])</span> it will be performed in synced transaction context. Synced transactions waits until all active replicas has committed the transaction (to disc) before returning from the mnesia:sync_transaction call. Using sync_transaction is useful for applications that are executing on several nodes and want to be sure that the update is performed on the remote nodes before a remote process is spawned or a message is sent to a remote process, and also when combining transaction writes with dirty_reads. This is also useful in situations where an application performs frequent or voluminous updates which may overload Mnesia on other nodes. </p> <p>By passing the same "fun" as argument to the function <span class="code">mnesia:async_dirty(Fun [, Args])</span> it will be performed in dirty context. The function calls will be mapped to the corresponding dirty functions. This will still involve logging, replication and subscriptions but there will be no locking, local transaction storage or commit protocols involved. Checkpoint retainers will be updated but will be updated "dirty". Thus, they will be updated asynchronously. The functions will wait for the operation to be performed on one node but not the others. If the table resides locally no waiting will occur. </p> <p>By passing the same "fun" as an argument to the function <span class="code">mnesia:sync_dirty(Fun [, Args])</span> it will be performed in almost the same context as <span class="code">mnesia:async_dirty/1,2</span>. The difference is that the operations are performed synchronously. The caller will wait for the updates to be performed on all active replicas. Using sync_dirty is useful for applications that are executing on several nodes and want to be sure that the update is performed on the remote nodes before a remote process is spawned or a message is sent to a remote process. This is also useful in situations where an application performs frequent or voluminous updates which may overload Mnesia on other nodes. </p> <p>You can check if your code is executed within a transaction with <span class="code">mnesia:is_transaction/0</span>, it returns <span class="code">true</span> when called inside a transaction context and false otherwise.</p> <p>Mnesia tables with storage type RAM_copies and disc_copies are implemented internally as "ets-tables" and it is possible for applications to access the these tables directly. This is only recommended if all options have been weighed and the possible outcomes are understood. By passing the earlier mentioned "fun" to the function <span class="code">mnesia:ets(Fun [, Args])</span> it will be performed but in a very raw context. The operations will be performed directly on the local ets tables assuming that the local storage type are RAM_copies and that the table is not replicated on other nodes. Subscriptions will not be triggered nor checkpoints updated, but this operation is blindingly fast. Disc resident tables should not be updated with the ets-function since the disc will not be updated. </p> <p>The Fun may also be passed as an argument to the function <span class="code">mnesia:activity/2,3,4</span> which enables usage of customized activity access callback modules. It can either be obtained directly by stating the module name as argument or implicitly by usage of the <span class="code">access_module</span> configuration parameter. A customized callback module may be used for several purposes, such as providing triggers, integrity constraints, run time statistics, or virtual tables. <br> The callback module does not have to access real Mnesia tables, it is free to do whatever it likes as long as the callback interface is fulfilled. <br> In Appendix C "The Activity Access Call Back Interface" the source code for one alternate implementation is provided (mnesia_frag.erl). The context sensitive function <span class="code">mnesia:table_info/2</span> may be used to provide virtual information about a table. One usage of this is to perform <span class="code">QLC</span> queries within an activity context with a customized callback module. By providing table information about table indices and other <span class="code">QLC</span> requirements, <span class="code">QLC</span> may be used as a generic query language to access virtual tables. </p> <p>QLC queries may be performed in all these activity contexts (transaction, sync_transaction, async_dirty, sync_dirty and ets). The ets activity will only work if the table has no indices. </p> <div class="note"> <div class="label">Note</div> <div class="content"><p> <p>The mnesia:dirty_* function always executes with async_dirty semantics regardless of which activity access contexts are invoked. They may even invoke contexts without any enclosing activity access context.</p> </p></div> </div> <h3><a name="id2272678">4.6 Nested transactions</a></h3> <p>Transactions may be nested in an arbitrary fashion. A child transaction must run in the same process as its parent. When a child transaction aborts, the caller of the child transaction will get the return value <span class="code">{aborted, Reason}</span> and any work performed by the child will be erased. If a child transaction commits, the records written by the child will be propagated to the parent. </p> <p>No locks are released when child transactions terminate. Locks created by a sequence of nested transactions are kept until the topmost transaction terminates. Furthermore, any updates performed by a nested transaction are only propagated in such a manner so that the parent of the nested transaction sees the updates. No final commitment will be done until the top level transaction is terminated. So, although a nested transaction returns <span class="code">{atomic, Val}</span>, if the enclosing parent transaction is aborted, the entire nested operation is aborted. </p> <p>The ability to have nested transaction with identical semantics as top level transaction makes it easier to write library functions that manipulate mnesia tables. </p> <p>Say for example that we have a function that adds a new subscriber to a telephony system:</p> <div class="example"><pre> add_subscriber(S) -> mnesia:transaction(fun() -> case mnesia:read( .......... </pre></div> <p>This function needs to be called as a transaction. Now assume that we wish to write a function that both calls the <span class="code">add_subscriber/1</span> function and is in itself protected by the context of a transaction. By simply calling the <span class="code">add_subscriber/1</span> from within another transaction, a nested transaction is created. </p> <p>It is also possible to mix different activity access contexts while nesting, but the dirty ones (async_dirty,sync_dirty and ets) will inherit the transaction semantics if they are called inside a transaction and thus it will grab locks and use two or three phase commit. </p> <div class="example"><pre> add_subscriber(S) -> mnesia:transaction(fun() -> %% Transaction context mnesia:read({some_tab, some_data}), mnesia:sync_dirty(fun() -> %% Still in a transaction context. case mnesia:read( ..) ..end), end). add_subscriber2(S) -> mnesia:sync_dirty(fun() -> %% In dirty context mnesia:read({some_tab, some_data}), mnesia:transaction(fun() -> %% In a transaction context. case mnesia:read( ..) ..end), end). </pre></div> <h3><a name="id2272772">4.7 Pattern Matching</a></h3> <a name="matching"></a> <p>When it is not possible to use <span class="code">mnesia:read/3</span> Mnesia provides the programmer with several functions for matching records against a pattern. The most useful functions of these are: </p> <div class="example"><pre> mnesia:select(Tab, MatchSpecification, LockKind) -> transaction abort | [ObjectList] mnesia:select(Tab, MatchSpecification, NObjects, Lock) -> transaction abort | {[Object],Continuation} | '$end_of_table' mnesia:select(Cont) -> transaction abort | {[Object],Continuation} | '$end_of_table' mnesia:match_object(Tab, Pattern, LockKind) -> transaction abort | RecordList </pre></div> <p>These functions matches a <span class="code">Pattern</span> against all records in table <span class="code">Tab</span>. In a <span class="code">mnesia:select</span> call <span class="code">Pattern</span> is a part of <span class="code">MatchSpecification</span> described below. It is not necessarily performed as an exhaustive search of the entire table. By utilizing indices and bound values in the key of the pattern, the actual work done by the function may be condensed into a few hash lookups. Using <span class="code">ordered_set</span> tables may reduce the search space if the keys are partially bound. </p> <p>The pattern provided to the functions must be a valid record, and the first element of the provided tuple must be the <span class="code">record_name</span> of the table. The special element <span class="code">'_'</span> matches any data structure in Erlang (also known as an Erlang term). The special elements <span class="code">'$<number>'</span> behaves as Erlang variables i.e. matches anything and binds the first occurrence and matches the coming occurrences of that variable against the bound value. </p> <p>Use the function <span class="code">mnesia:table_info(Tab, wild_pattern)</span> to obtain a basic pattern which matches all records in a table or use the default value in record creation. Do not make the pattern hard coded since it will make your code more vulnerable to future changes of the record definition. </p> <div class="example"><pre> Wildpattern = mnesia:table_info(employee, wild_pattern), %% Or use Wildpattern = #employee{_ = '_'}, </pre></div> <p>For the employee table the wild pattern will look like:</p> <div class="example"><pre> {employee, '_', '_', '_', '_', '_',' _'}. </pre></div> <p>In order to constrain the match you must replace some of the <span class="code">'_'</span> elements. The code for matching out all female employees, looks like: </p> <div class="example"><pre> Pat = #employee{sex = female, _ = '_'}, F = fun() -> mnesia:match_object(Pat) end, Females = mnesia:transaction(F). </pre></div> <p>It is also possible to use the match function if we want to check the equality of different attributes. Assume that we want to find all employees which happens to have a employee number which is equal to their room number: </p> <div class="example"><pre> Pat = #employee{emp_no = '$1', room_no = '$1', _ = '_'}, F = fun() -> mnesia:match_object(Pat) end, Odd = mnesia:transaction(F). </pre></div> <p>The function <span class="code">mnesia:match_object/3</span> lacks some important features that <span class="code">mnesia:select/3</span> have. For example <span class="code">mnesia:match_object/3</span> can only return the matching records, and it can not express constraints other then equality. If we want to find the names of the male employees on the second floor we could write: </p> <div class="example"><pre> MatchHead = #employee{name='$1', sex=male, room_no={'$2', '_'}, _='_'}, Guard = [{'>=', '$2', 220},{'<', '$2', 230}], Result = '$1', mnesia:select(employee,[{MatchHead, Guard, [Result]}])</pre></div> <p>Select can be used to add additional constraints and create output which can not be done with <span class="code">mnesia:match_object/3</span>. </p> <p>The second argument to select is a <span class="code">MatchSpecification</span>. A <span class="code">MatchSpecification</span> is list of <span class="code">MatchFunctions</span>, where each <span class="code">MatchFunction</span> consists of a tuple containing <span class="code">{MatchHead, MatchCondition, MatchBody}</span>. <span class="code">MatchHead</span> is the same pattern used in <span class="code">mnesia:match_object/3</span> described above. <span class="code">MatchCondition</span> is a list of additional constraints applied to each record, and <span class="code">MatchBody</span> is used to construct the return values. </p> <p>A detailed explanation of match specifications can be found in the <strong>Erts users guide: Match specifications in Erlang </strong>, and the ets/dets documentations may provide some additional information. </p> <p>The functions <span class="code">select/4</span> and <span class="code">select/1</span> are used to get a limited number of results, where the <span class="code">Continuation</span> are used to get the next chunk of results. Mnesia uses the <span class="code">NObjects</span> as an recommendation only, thus more or less results then specified with <span class="code">NObjects</span> may be returned in the result list, even the empty list may be returned despite there are more results to collect. </p> <div class="warning"> <div class="label">Warning</div> <div class="content"><p> <p>There is a severe performance penalty in using <span class="code">mnesia:select/[1|2|3|4]</span> after any modifying operations are done on that table in the same transaction, i.e. avoid using <span class="code">mnesia:write/1</span> or <span class="code">mnesia:delete/1</span> before a <span class="code">mnesia:select</span> in the same transaction.</p> </p></div> </div> <p>If the key attribute is bound in a pattern, the match operation is very efficient. However, if the key attribute in a pattern is given as <span class="code">'_'</span>, or <span class="code">'$1'</span>, the whole <span class="code">employee</span> table must be searched for records that match. Hence if the table is large, this can become a time consuming operation, but it can be remedied with indices (refer to Chapter 5: <span class="bold_code"><a href="Mnesia_chap5.html#indexing">Indexing</a></span>) if <span class="code">mnesia:match_object</span> is used. </p> <p>QLC queries can also be used to search Mnesia tables. By using <span class="code">mnesia:table/[1|2]</span> as the generator inside a QLC query you let the query operate on a mnesia table. Mnesia specific options to <span class="code">mnesia:table/2</span> are {lock, Lock}, {n_objects,Integer} and {traverse, SelMethod}. The <span class="code">lock</span> option specifies whether mnesia should acquire a read or write lock on the table, and <span class="code">n_objects</span> specifies how many results should be returned in each chunk to QLC. The last option is <span class="code">traverse</span> and it specifies which function mnesia should use to traverse the table. Default <span class="code">select</span> is used, but by using <span class="code">{traverse, {select, MatchSpecification}}</span> as an option to <span class="code">mnesia:table/2</span> the user can specify it's own view of the table. </p> <p>If no options are specified a read lock will acquired and 100 results will be returned in each chunk, and select will be used to traverse the table, i.e.: </p> <div class="example"><pre> mnesia:table(Tab) -> mnesia:table(Tab, [{n_objects,100},{lock, read}, {traverse, select}]). </pre></div> <p>The function <span class="code">mnesia:all_keys(Tab)</span> returns all keys in a table.</p> <h3><a name="id2273172">4.8 Iteration</a></h3> <a name="iteration"></a> <p>Mnesia provides a couple of functions which iterates over all the records in a table. </p> <div class="example"><pre> mnesia:foldl(Fun, Acc0, Tab) -> NewAcc | transaction abort mnesia:foldr(Fun, Acc0, Tab) -> NewAcc | transaction abort mnesia:foldl(Fun, Acc0, Tab, LockType) -> NewAcc | transaction abort mnesia:foldr(Fun, Acc0, Tab, LockType) -> NewAcc | transaction abort </pre></div> <p>These functions iterate over the mnesia table <span class="code">Tab</span> and apply the function <span class="code">Fun</span> to each record. The <span class="code">Fun</span> takes two arguments, the first argument is a record from the table and the second argument is the accumulator. The <span class="code">Fun</span> return a new accumulator. </p> <p>The first time the <span class="code">Fun</span> is applied <span class="code">Acc0</span> will be the second argument. The next time the <span class="code">Fun</span> is called the return value from the previous call, will be used as the second argument. The term the last call to the Fun returns will be the return value of the <span class="code">fold[lr]</span> function. </p> <p>The difference between <span class="code">foldl</span> and <span class="code">foldr</span> is the order the table is accessed for <span class="code">ordered_set</span> tables, for every other table type the functions are equivalent. </p> <p><span class="code">LockType</span> specifies what type of lock that shall be acquired for the iteration, default is <span class="code">read</span>. If records are written or deleted during the iteration a write lock should be acquired. </p> <p>These functions might be used to find records in a table when it is impossible to write constraints for <span class="code">mnesia:match_object/3</span>, or when you want to perform some action on certain records. </p> <p>For example finding all the employees who has a salary below 10 could look like:</p> <div class="example"><pre> find_low_salaries() -> Constraint = fun(Emp, Acc) when Emp#employee.salary < 10 -> [Emp | Acc]; (_, Acc) -> Acc end, Find = fun() -> mnesia:foldl(Constraint, [], employee) end, mnesia:transaction(Find). </pre></div> <p>Raising the salary to 10 for everyone with a salary below 10 and return the sum of all raises:</p> <div class="example"><pre> increase_low_salaries() -> Increase = fun(Emp, Acc) when Emp#employee.salary < 10 -> OldS = Emp#employee.salary, ok = mnesia:write(Emp#employee{salary = 10}), Acc + 10 - OldS; (_, Acc) -> Acc end, IncLow = fun() -> mnesia:foldl(Increase, 0, employee, write) end, mnesia:transaction(IncLow). </pre></div> <p>A lot of nice things can be done with the iterator functions but some caution should be taken about performance and memory utilization for large tables. </p> <p>Call these iteration functions on nodes that contain a replica of the table. Each call to the function <span class="code">Fun</span> access the table and if the table resides on another node it will generate a lot of unnecessary network traffic. </p> <p>Mnesia also provides some functions that make it possible for the user to iterate over the table. The order of the iteration is unspecified if the table is not of the <span class="code">ordered_set</span> type. </p> <div class="example"><pre> mnesia:first(Tab) -> Key | transaction abort mnesia:last(Tab) -> Key | transaction abort mnesia:next(Tab,Key) -> Key | transaction abort mnesia:prev(Tab,Key) -> Key | transaction abort mnesia:snmp_get_next_index(Tab,Index) -> {ok, NextIndex} | endOfTable </pre></div> <p>The order of first/last and next/prev are only valid for <span class="code">ordered_set</span> tables, for all other tables, they are synonyms. When the end of the table is reached the special key <span class="code">'$end_of_table'</span> is returned.</p> <p>If records are written and deleted during the traversal, use <span class="code">mnesia:fold[lr]/4</span> with a <span class="code">write</span> lock. Or <span class="code">mnesia:write_lock_table/1</span> when using first and next.</p> <p>Writing or deleting in transaction context creates a local copy of each modified record, so modifying each record in a large table uses a lot of memory. Mnesia will compensate for every written or deleted record during the iteration in a transaction context, which may reduce the performance. If possible avoid writing or deleting records in the same transaction before iterating over the table.</p> <p>In dirty context, i.e. <span class="code">sync_dirty</span> or <span class="code">async_dirty</span>, the modified records are not stored in a local copy; instead, each record is updated separately. This generates a lot of network traffic if the table has a replica on another node and has all the other drawbacks that dirty operations have. Especially for the <span class="code">mnesia:first/1</span> and <span class="code">mnesia:next/2</span> commands, the same drawbacks as described above for <span class="code">dirty_first</span> and <span class="code">dirty_next</span> applies, i.e. no writes to the table should be done during iteration.</p> <p></p> </div> <div class="footer"> <hr> <p>Copyright © 1997-2010 Ericsson AB. All Rights Reserved.</p> </div> </div> </div></body> </html>