ref/transapp/throughput.html

<!--$Id: throughput.so,v 10.31 2002/04/02 17:07:05 bostic Exp $-->
<!--Copyright (c) 1997,2008 Oracle.  All rights reserved.-->
<!--See the file LICENSE for redistribution information.-->
<html>
<head>
<title>Berkeley DB Reference Guide: Transaction throughput</title>
<meta name="description" content="Berkeley DB: An embedded database programmatic toolkit.">
<meta name="keywords" content="embedded,database,programmatic,toolkit,btree,hash,hashing,transaction,transactions,locking,logging,access method,access methods,Java,C,C++">
</head>
<body bgcolor=white>
<a name="2"><!--meow--></a>
<table width="100%"><tr valign=top>
<td><b><dl><dt>Berkeley DB Reference Guide:<dd>Berkeley DB Transactional Data Store Applications</dl></b></td>
<td align=right><a href="../transapp/tune.html"><img src="../../images/prev.gif" alt="Prev"></a><a href="../toc.html"><img src="../../images/ref.gif" alt="Ref"></a><a href="../transapp/faq.html"><img src="../../images/next.gif" alt="Next"></a>
</td></tr></table>
<p align=center><b>Transaction throughput</b></p>
<p>Generally, the speed of a database system is measured by the
<i>transaction throughput</i>, expressed as a number of
transactions per second.  The two gating factors for Berkeley DB performance
in a transactional system are usually the underlying database files and
the log file.  Both are factors because they require disk I/O, which is
slow relative to other system resources such as CPU.</p>
<p>In the worst-case scenario:</p>
<p><ul type=disc>
<li>Database access is truly random and the database is too large for any
significant percentage of it to fit into the cache, resulting in a
single I/O per requested key/data pair.
<li>Both the database and the log are on a single disk.
</ul>
<p>This means that for each transaction, Berkeley DB is potentially performing
several filesystem operations:</p>
<p><ul type=disc>
<li>Disk seek to database file
<li>Database file read
<li>Disk seek to log file
<li>Log file write
<li>Flush log file information to disk
<li>Disk seek to update log file metadata (for example, inode information)
<li>Log metadata write
<li>Flush log file metadata to disk
</ul>
<p>There are a number of ways to increase transactional throughput, all of
which attempt to decrease the number of filesystem operations per
transaction.  First, the Berkeley DB software includes support for
<i>group commit</i>.  Group commit simply means that when the
information about one transaction is flushed to disk, the information
for any other waiting transactions will be flushed to disk at the same
time, potentially amortizing a single log write over a large number of
transactions.  There are additional tuning parameters which may be
useful to application writers:</p>
<p><ul type=disc>
<li>Tune the size of the database cache.  If the Berkeley DB key/data pairs used
during the transaction are found in the database cache, the seek and read
from the database are no longer necessary, resulting in two fewer
filesystem operations per transaction.  To determine whether your cache
size is too small, see <a href="../../ref/am_conf/cachesize.html">Selecting
a cache size</a>.
<li>Put the database and the log files on different disks.  This allows reads
and writes to the log files and the database files to be performed
concurrently.
<li>Set the filesystem configuration so that file access and modification times
are not updated.  Note that although the file access and modification times
are not used by Berkeley DB, this may affect other programs -- so be careful.
<li>Upgrade your hardware.  When considering the hardware on which to run your
application, however, it is important to consider the entire system.  The
controller and bus can have as much to do with the disk performance as
the disk itself.  It is also important to remember that throughput is
rarely the limiting factor, and that disk seek times are normally the true
performance issue for Berkeley DB.
<li>Turn on the <a href="../../api_c/env_set_flags.html#DB_TXN_WRITE_NOSYNC">DB_TXN_WRITE_NOSYNC</a> or <a href="../../api_c/env_set_flags.html#DB_TXN_NOSYNC">DB_TXN_NOSYNC</a> flags.
This changes the Berkeley DB behavior so that the log files are not written
and/or flushed when transactions are committed.  Although this change
will greatly increase your transaction throughput, it means that
transactions will exhibit the ACI (atomicity, consistency, and
isolation) properties, but not D (durability).  Database integrity will
be maintained, but it is possible that some number of the most recently
committed transactions may be undone during recovery instead of being
redone.
</ul>
<p>If you are bottlenecked on logging, the following test will help you
confirm that the number of transactions per second that your application
does is reasonable for the hardware on which you're running.  Your test
program should repeatedly perform the following operations:</p>
<p><ul type=disc>
<li>Seek to the beginning of a file
<li>Write to the file
<li>Flush the file write to disk
</ul>
<p>The number of times that you can perform these three operations per
second is a rough measure of the minimum number of transactions per
second of which the hardware is capable.  This test simulates the
operations applied to the log file. (As a simplifying assumption in this
experiment, we assume that the database files are either on a separate
disk; or that they fit, with some few exceptions, into the database
cache.)  We do not have to directly simulate updating the log file
directory information because it will normally be updated and flushed
to disk as a result of flushing the log file write to disk.</p>
<p>Running this test program, in which we write 256 bytes for 1000 operations
on reasonably standard commodity hardware (Pentium II CPU, SCSI disk),
returned the following results:</p>
<blockquote><pre>% testfile -b256 -o1000
running: 1000 ops
Elapsed time: 16.641934 seconds
1000 ops:   60.09 ops per second</pre></blockquote>
<p>Note that the number of bytes being written to the log as part of each
transaction can dramatically affect the transaction throughput.  The
test run used 256, which is a reasonable size log write.  Your log
writes may be different.  To determine your average log write size, use
the <a href="../../utility/db_stat.html">db_stat</a> utility to display your log statistics.</p>
<p>As a quick sanity check, the average seek time is 9.4 msec for this
particular disk, and the average latency is 4.17 msec.  That results in
a minimum requirement for a data transfer to the disk of 13.57 msec, or
a maximum of 74 transfers per second.  This is close enough to the
previous 60 operations per second (which wasn't done on a quiescent
disk) that the number is believable.</p>
<p>An implementation of the previous <a href="writetest.cs">example test
program</a> for IEEE/ANSI Std 1003.1 (POSIX) standard systems is included in the Berkeley DB
distribution.</p>
<table width="100%"><tr><td><br></td><td align=right><a href="../transapp/tune.html"><img src="../../images/prev.gif" alt="Prev"></a><a href="../toc.html"><img src="../../images/ref.gif" alt="Ref"></a><a href="../transapp/faq.html"><img src="../../images/next.gif" alt="Next"></a>
</td></tr></table>
<p><font size=1>Copyright (c) 1996,2008 Oracle.  All rights reserved.</font>
</body>
</html>