backuprestore.html [plain text]
<?xml version="1.0" encoding="UTF-8" standalone="no"?>
<!DOCTYPE html PUBLIC "-//W3C//DTD XHTML 1.0 Transitional//EN" "http://www.w3.org/TR/xhtml1/DTD/xhtml1-transitional.dtd">
<html xmlns="http://www.w3.org/1999/xhtml">
<head>
<meta http-equiv="Content-Type" content="text/html; charset=UTF-8" />
<title>Backup Procedures</title>
<link rel="stylesheet" href="gettingStarted.css" type="text/css" />
<meta name="generator" content="DocBook XSL Stylesheets V1.62.4" />
<link rel="home" href="index.html" title="Getting Started with Berkeley DB Transaction Processing" />
<link rel="up" href="filemanagement.html" title="Chapter 5. Managing DB Files" />
<link rel="previous" href="filemanagement.html" title="Chapter 5. Managing DB Files" />
<link rel="next" href="recovery.html" title="Recovery Procedures" />
</head>
<body>
<div class="navheader">
<table width="100%" summary="Navigation header">
<tr>
<th colspan="3" align="center">Backup Procedures</th>
</tr>
<tr>
<td width="20%" align="left"><a accesskey="p" href="filemanagement.html">Prev</a> </td>
<th width="60%" align="center">Chapter 5. Managing DB Files</th>
<td width="20%" align="right"> <a accesskey="n" href="recovery.html">Next</a></td>
</tr>
</table>
<hr />
</div>
<div class="sect1" lang="en" xml:lang="en">
<div class="titlepage">
<div>
<div>
<h2 class="title" style="clear: both"><a id="backuprestore"></a>Backup Procedures</h2>
</div>
</div>
<div></div>
</div>
<p>
<span class="emphasis"><em>Durability</em></span> is an important part of your
transactional guarantees. It means that once a transaction has been
successfully committed, your application will always see the results of that
transaction.
</p>
<p>
Of course, no software algorithm can guarantee durability in the face of physical data loss. Hard drives
can fail, and if you have not copied your data to locations other than your primary disk drives,
then you will lose data when those drives fail. Therefore, in order to truly obtain a durability
guarantee, you need to ensure that any data stored on disk is backed up to secondary or alternative storage,
such as secondary disk drives, or offline tapes.
</p>
<p>
There are three different types of backups that you can
perform with DB databases and log files. They are:
</p>
<div class="itemizedlist">
<ul type="disc">
<li>
<p>
Offline backups
</p>
<p>
This type of backup is perhaps the easiest to perform as it
involves simply copying database and log files to an
offline storage area. It also gives you a snapshot of the
database at a fixed, known point in time. However, you
cannot perform this type of a backup while you are performing
writes to the database.
</p>
</li>
<li>
<p>
Hot backups
</p>
<p>
This type of backup gives you a snapshot of your database.
Since your application can be writing to the database at the time that the
snapshot is being taken, you do not necessarily know what
the exact state of the database is for that given snapshot.
</p>
</li>
<li>
<p>
Incremental backups
</p>
<p>
This type of backup refreshes a previously performed backup.
</p>
</li>
</ul>
</div>
<p>
Once you have performed a backup, you can
perform <span class="emphasis"><em>catastrophic recovery</em></span> to restore
your databases from the backup. See
<a href="recovery.html#catastrophicrecovery">Catastrophic Recovery</a>
for more information.
</p>
<p>
Note that you can also maintain a hot failover. See
<a href="hotfailover.html">Using Hot Failovers</a>
for more information.
</p>
<div class="sect2" lang="en" xml:lang="en">
<div class="titlepage">
<div>
<div>
<h3 class="title"><a id="copyutilities"></a>About Unix Copy Utilities</h3>
</div>
</div>
<div></div>
</div>
<p>
If you are copying database files you must copy databases atomically,
in multiples of the database page size. In other words, the reads made by
the copy program must not be interleaved with writes by
other threads of control, and the copy program must read the
databases in multiples of the underlying database page size.
Generally, this is not a problem because operating systems
already make this guarantee and system utilities normally
read in power-of-2 sized chunks, which are larger than the
largest possible Berkeley DB database page size.
</p>
<p>
On some platforms (most notably, some releases of Solaris), the copy utility (<tt class="literal">cp</tt>) was
implemented using the <tt class="function">mmap()</tt> system call rather than the
<tt class="function">read()</tt> system call. Because <tt class="function">mmap()</tt> did not make the same
guarantee of read atomicity as did <tt class="function">read()</tt>, the <tt class="literal">cp</tt> utility
could create corrupted copies of the databases.
</p>
<p>
Also, some platforms have implementations of the <tt class="literal">tar</tt> utility that performs 10KB block
reads by default. Even when an output block size is specified, the utility will still not read the
underlying databases in multiples of the specified block size. Again, the result can be a corrupted backup.
</p>
<p>
To fix these problems, use the <tt class="literal">dd</tt> utility instead of <tt class="literal">cp</tt> or
<tt class="literal">tar</tt>. When you use <tt class="literal">dd</tt>, make sure you specify a block size that is
equal to, or an even multiple of, your database page size. Finally, if you plan to use a system
utility to copy database files, you may want to use a system call trace utility (for example,
<tt class="literal">ktrace</tt> or <tt class="literal">truss</tt>) to make sure you are not using a I/O size that is
smaller than your database page size. You can also use these utilities to make sure the system utility is
not using a system call other than <tt class="function">read()</tt>.
</p>
</div>
<div class="sect2" lang="en" xml:lang="en">
<div class="titlepage">
<div>
<div>
<h3 class="title"><a id="standardbackup"></a>Offline Backups</h3>
</div>
</div>
<div></div>
</div>
<p>
To create an offline backup:
</p>
<div class="orderedlist">
<ol type="1">
<li>
<p>
Commit or abort all on-going transactions.
</p>
</li>
<li>
<p>
Pause all database writes.
</p>
</li>
<li>
<p>
Force a checkpoint. See
<a href="filemanagement.html#checkpoints">Checkpoints</a>
for details.
</p>
</li>
<li>
<p>
Copy all your database files to the backup location.
<span>
Note that you can simply copy all of the database
files, or you can determine which database files
have been written during the lifetime of the current
logs. To do this, use either the
<span>
<tt class="methodname">DbEnv::log_archive()</tt>
method with the <tt class="literal">DB_ARCH_DATA</tt>
option,
</span>
or use the <span><b class="command">db_archive</b></span>
command with the <tt class="literal">-s</tt> option.
</span>
</p>
<p>
However, be aware that backing up just the modified databases only works if you have all of your
log files. If you have been removing log files for any reason then using
<span>
<tt class="methodname">log_archive()</tt>
</span>
can result in an
unrecoverable backup because you might not be notified of a database file that was modified.
</p>
</li>
<li>
<p>
Copy the <span class="emphasis"><em>last</em></span> log file to your backup location.
Your log files are named
<tt class="literal">log.<span class="emphasis"><em>xxxxxxxxxx</em></span></tt>,
where <span class="emphasis"><em>xxxxxxxxxx</em></span> is a
sequential number. The last log file is the file
with the highest number.
</p>
</li>
</ol>
</div>
</div>
<div class="sect2" lang="en" xml:lang="en">
<div class="titlepage">
<div>
<div>
<h3 class="title"><a id="hotbackup"></a>Hot Backup</h3>
</div>
</div>
<div></div>
</div>
<p>
To create a hot backup, you do not have to stop database
operations. Transactions may be on-going and you can be writing
to your database at the time of the backup. However, this means
that you do not know exactly what the state of your database is
at the time of the backup.
</p>
<p>
You can use the <span><b class="command">db_hotbackup</b></span> command line utility to create a hot backup for you. This
utility will (optionally) run a checkpoint and the copy all necessary files to a target directory.
</p>
<p>
Alternatively, you can manually create a hot backup as follows:
</p>
<div class="orderedlist">
<ol type="1">
<li>
<p>
Copy all your database files to the backup location.
<span>
Note that you can simply copy all of the database
files, or you can determine which database files
have been written during the lifetime of the current
logs. To do this, use either the
<span>
<tt class="methodname">DbEnv::log_archive()</tt>
with the <tt class="literal">DB_ARCH_DATA</tt>
option,
</span>
or use the <span><b class="command">db_archive</b></span>
command with the <tt class="literal">-s</tt> option.
</span>
</p>
</li>
<li>
<p>
Copy all logs to your backup location.
</p>
</li>
</ol>
</div>
<div class="note" style="margin-left: 0.5in; margin-right: 0.5in;">
<h3 class="title">Note</h3>
<p>
It is important to copy your database files <span class="emphasis"><em>and
then</em></span> your logs. In this way,
you can complete or roll back any database operations that were only partially completed
when you copied the databases.
</p>
</div>
</div>
<div class="sect2" lang="en" xml:lang="en">
<div class="titlepage">
<div>
<div>
<h3 class="title"><a id="incrementalbackups"></a>Incremental Backups</h3>
</div>
</div>
<div></div>
</div>
<p>
Once you have created a full backup (that is, either a
offline or hot backup), you can create incremental backups.
To do this, simply copy all of your currently existing log
files to your backup location.
</p>
<p>
Incremental backups do not require you to run a checkpoint
or to cease database write operations.
</p>
<p>
When you are working with incremental backups, remember
that the greater the number of log files contained in
your backup, the longer recovery will take.
You should run full backups
on some interval, and then do incremental backups on a shorter interval.
How frequently you need to run a full backup
is determined by the rate at which your databases change and
how sensitive your application is to lengthy recoveries
(should one be required).
</p>
<p>
You can also shorten recovery time by running recovery against the backup as you take each incremental
backup. Running recovery as you go means that there will be less work for DB to do if you should
ever need to restore your environment from the backup.
</p>
</div>
</div>
<div class="navfooter">
<hr />
<table width="100%" summary="Navigation footer">
<tr>
<td width="40%" align="left"><a accesskey="p" href="filemanagement.html">Prev</a> </td>
<td width="20%" align="center">
<a accesskey="u" href="filemanagement.html">Up</a>
</td>
<td width="40%" align="right"> <a accesskey="n" href="recovery.html">Next</a></td>
</tr>
<tr>
<td width="40%" align="left" valign="top">Chapter 5. Managing DB Files </td>
<td width="20%" align="center">
<a accesskey="h" href="index.html">Home</a>
</td>
<td width="40%" align="right" valign="top"> Recovery Procedures</td>
</tr>
</table>
</div>
</body>
</html>