1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 188 189 190 191 192 193 194 195 196 197 198 199 200 201 202 203 204 205 206 207 208 209 210 211 212 213 214 215 216 217 218 219 220 221 222 223 224 225 226 227 228 229 230 231 232 233 234 235 236 237 238 239 240 241 242 243 244 245 246 247 248 249 250 251 252 253 254 255 256 257 258 259 260 261 262 263 264 265 266 267 268 269 270 271 272 273 274 275 276 277 278 279 280 281 282 283 284 285 286 287 288 289 290 291 292 293 294 295 296 297 298 299 300 301 302 303 304 305 306 307 308 309 310 311 312 313 314 315 316 317 318 319 320 321 322 323 324 325 326 327 328 329 330 331 332 333 334 335 336
|
<?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>General access method configuration</title>
<link rel="stylesheet" href="gettingStarted.css" type="text/css" />
<meta name="generator" content="DocBook XSL Stylesheets V1.73.2" />
<link rel="start" href="index.html" title="Berkeley DB Programmer's Reference Guide" />
<link rel="up" href="am_conf.html" title="Chapter 2. Access Method Configuration" />
<link rel="prev" href="am_conf_logrec.html" title="Logical record numbers" />
<link rel="next" href="bt_conf.html" title="Btree access method specific configuration" />
</head>
<body>
<div class="navheader">
<table width="100%" summary="Navigation header">
<tr>
<th colspan="3" align="center">General access method configuration</th>
</tr>
<tr>
<td width="20%" align="left"><a accesskey="p" href="am_conf_logrec.html">Prev</a> </td>
<th width="60%" align="center">Chapter 2.
Access Method Configuration
</th>
<td width="20%" align="right"> <a accesskey="n" href="bt_conf.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="general_am_conf"></a>General access method configuration</h2>
</div>
</div>
</div>
<div class="toc">
<dl>
<dt>
<span class="sect2">
<a href="general_am_conf.html#am_conf_pagesize">Selecting a page size</a>
</span>
</dt>
<dt>
<span class="sect2">
<a href="general_am_conf.html#am_conf_cachesize">Selecting a cache size</a>
</span>
</dt>
<dt>
<span class="sect2">
<a href="general_am_conf.html#am_conf_byteorder">Selecting a byte order</a>
</span>
</dt>
<dt>
<span class="sect2">
<a href="general_am_conf.html#am_conf_dup">Duplicate data items</a>
</span>
</dt>
<dt>
<span class="sect2">
<a href="general_am_conf.html#am_conf_malloc">Non-local memory allocation</a>
</span>
</dt>
</dl>
</div>
<p>
There are a series of configuration tasks which are common to all
access methods. They are described in the following sections.
</p>
<div class="sect2" lang="en" xml:lang="en">
<div class="titlepage">
<div>
<div>
<h3 class="title"><a id="am_conf_pagesize"></a>Selecting a page size</h3>
</div>
</div>
</div>
<p>The size of the pages used in the underlying database can be specified by
calling the <a href="../api_reference/C/dbset_pagesize.html" class="olink">DB->set_pagesize()</a> method. The minimum page size is 512 bytes
and the maximum page size is 64K bytes, and must be a power of two. If
no page size is specified by the application, a page size is selected
based on the underlying filesystem I/O block size. (A page size selected
in this way has a lower limit of 512 bytes and an upper limit of 16K
bytes.)</p>
<p>There are several issues to consider when selecting a pagesize: overflow
record sizes, locking, I/O efficiency, and recoverability.</p>
<p>First, the page size implicitly sets the size of an overflow record.
Overflow records are key or data items that are too large to fit on a
normal database page because of their size, and are therefore stored in
overflow pages. Overflow pages are pages that exist outside of the normal
database structure. For this reason, there is often a significant
performance penalty associated with retrieving or modifying overflow
records. Selecting a page size that is too small, and which forces the
creation of large numbers of overflow pages, can seriously impact the
performance of an application.</p>
<p>Second, in the Btree, Hash and Recno access methods, the finest-grained
lock that Berkeley DB acquires is for a page. (The Queue access method
generally acquires record-level locks rather than page-level locks.)
Selecting a page size that is too large, and which causes threads or
processes to wait because other threads of control are accessing or
modifying records on the same page, can impact the performance of your
application.</p>
<p>Third, the page size specifies the granularity of I/O from the database
to the operating system. Berkeley DB will give a page-sized unit of bytes to
the operating system to be scheduled for reading/writing from/to the
disk. For many operating systems, there is an internal <span class="bold"><strong>block
size</strong></span> which is used as the granularity of I/O from the operating system
to the disk. Generally, it will be more efficient for Berkeley DB to write
filesystem-sized blocks to the operating system and for the operating
system to write those same blocks to the disk.</p>
<p>Selecting a database page size smaller than the filesystem block size
may cause the operating system to coalesce or otherwise manipulate Berkeley DB
pages and can impact the performance of your application. When the page
size is smaller than the filesystem block size and a page written by
Berkeley DB is not found in the operating system's cache, the operating system
may be forced to read a block from the disk, copy the page into the
block it read, and then write out the block to disk, rather than simply
writing the page to disk. Additionally, as the operating system is
reading more data into its buffer cache than is strictly necessary to
satisfy each Berkeley DB request for a page, the operating system buffer cache
may be wasting memory.</p>
<p>Alternatively, selecting a page size larger than the filesystem block
size may cause the operating system to read more data than necessary.
On some systems, reading filesystem blocks sequentially may cause the
operating system to begin performing read-ahead. If requesting a single
database page implies reading enough filesystem blocks to satisfy the
operating system's criteria for read-ahead, the operating system may do
more I/O than is required.</p>
<p>Fourth, when using the Berkeley DB Transactional Data Store product, the page size may affect the errors
from which your database can recover See
<a class="xref" href="transapp_reclimit.html" title="Berkeley DB recoverability">Berkeley DB recoverability</a> for more
information.</p>
</div>
<div class="sect2" lang="en" xml:lang="en">
<div class="titlepage">
<div>
<div>
<h3 class="title"><a id="am_conf_cachesize"></a>Selecting a cache size</h3>
</div>
</div>
</div>
<p>The size of the cache used for the underlying database can be specified
by calling the <a href="../api_reference/C/dbset_cachesize.html" class="olink">DB->set_cachesize()</a> method.
Choosing a cache size is, unfortunately, an art. Your cache must be at
least large enough for your working set plus some overlap for unexpected
situations.</p>
<p>When using the Btree access method, you must have a cache big enough for
the minimum working set for a single access. This will include a root
page, one or more internal pages (depending on the depth of your tree),
and a leaf page. If your cache is any smaller than that, each new page
will force out the least-recently-used page, and Berkeley DB will re-read the
root page of the tree anew on each database request.</p>
<p>If your keys are of moderate size (a few tens of bytes) and your pages
are on the order of 4KB to 8KB, most Btree applications will be only
three levels. For example, using 20 byte keys with 20 bytes of data
associated with each key, a 8KB page can hold roughly 400 keys (or 200
key/data pairs), so a fully populated three-level Btree will hold 32
million key/data pairs, and a tree with only a 50% page-fill factor will
still hold 16 million key/data pairs. We rarely expect trees to exceed
five levels, although Berkeley DB will support trees up to 255 levels.</p>
<p>The rule-of-thumb is that cache is good, and more cache is better.
Generally, applications benefit from increasing the cache size up to a
point, at which the performance will stop improving as the cache size
increases. When this point is reached, one of two things have happened:
either the cache is large enough that the application is almost never
having to retrieve information from disk, or, your application is doing
truly random accesses, and therefore increasing size of the cache doesn't
significantly increase the odds of finding the next requested information
in the cache. The latter is fairly rare -- almost all applications show
some form of locality of reference.</p>
<p>That said, it is important not to increase your cache size beyond the
capabilities of your system, as that will result in reduced performance.
Under many operating systems, tying down enough virtual memory will cause
your memory and potentially your program to be swapped. This is
especially likely on systems without unified OS buffer caches and virtual
memory spaces, as the buffer cache was allocated at boot time and so
cannot be adjusted based on application requests for large amounts of
virtual memory.</p>
<p>For example, even if accesses are truly random within a Btree, your
access pattern will favor internal pages to leaf pages, so your cache
should be large enough to hold all internal pages. In the steady state,
this requires at most one I/O per operation to retrieve the appropriate
leaf page.</p>
<p>You can use the <a href="../api_reference/C/db_stat.html" class="olink">db_stat utility</a> to monitor the effectiveness of
your cache. The following output is excerpted from the output of that
utility's <span class="bold"><strong>-m</strong></span> option:</p>
<pre class="programlisting">prompt: db_stat -m
131072 Cache size (128K).
4273 Requested pages found in the cache (97%).
134 Requested pages not found in the cache.
18 Pages created in the cache.
116 Pages read into the cache.
93 Pages written from the cache to the backing file.
5 Clean pages forced from the cache.
13 Dirty pages forced from the cache.
0 Dirty buffers written by trickle-sync thread.
130 Current clean buffer count.
4 Current dirty buffer count.
</pre>
<p>The statistics for this cache say that there have been 4,273 requests of
the cache, and only 116 of those requests required an I/O from disk. This
means that the cache is working well, yielding a 97% cache hit rate. The
<a href="../api_reference/C/db_stat.html" class="olink">db_stat utility</a> will present these statistics both for the cache
as a whole and for each file within the cache separately.</p>
</div>
<div class="sect2" lang="en" xml:lang="en">
<div class="titlepage">
<div>
<div>
<h3 class="title"><a id="am_conf_byteorder"></a>Selecting a byte order</h3>
</div>
</div>
</div>
<p>Database files created by Berkeley DB can be created in either little- or
big-endian formats. The byte order used for the underlying database
is specified by calling the <a href="../api_reference/C/dbset_lorder.html" class="olink">DB->set_lorder()</a> method. If no order
is selected, the native format of the machine on which the database is
created will be used.</p>
<p>Berkeley DB databases are architecture independent, and any format database can
be used on a machine with a different native format. In this case, as
each page that is read into or written from the cache must be converted
to or from the host format, and databases with non-native formats will
incur a performance penalty for the run-time conversion.</p>
<p>
<span class="bold">
<strong>It is important to note that the Berkeley DB access methods do no data
conversion for application specified data. Key/data pairs written on a
little-endian format architecture will be returned to the application
exactly as they were written when retrieved on a big-endian format
architecture.</strong>
</span>
</p>
</div>
<div class="sect2" lang="en" xml:lang="en">
<div class="titlepage">
<div>
<div>
<h3 class="title"><a id="am_conf_dup"></a>Duplicate data items</h3>
</div>
</div>
</div>
<p>The Btree and Hash access methods support the creation of multiple data
items for a single key item. By default, multiple data items are not
permitted, and each database store operation will overwrite any previous
data item for that key. To configure Berkeley DB for duplicate data items,
call the <a href="../api_reference/C/dbset_flags.html" class="olink">DB->set_flags()</a> method with the <a href="../api_reference/C/dbset_flags.html#dbset_flags_DB_DUP" class="olink">DB_DUP</a> flag. Only one
copy of the key will be stored for each set of duplicate data items.
If the Btree access method comparison routine returns that two keys
compare equally, it is undefined which of the two keys will be stored
and returned from future database operations.</p>
<p>By default, Berkeley DB stores duplicates in the order in which they were added,
that is, each new duplicate data item will be stored after any already
existing data items. This default behavior can be overridden by using
the <a href="../api_reference/C/dbcput.html" class="olink">DBC->put()</a> method and one of the <a href="../api_reference/C/dbcput.html#put_DB_AFTER" class="olink">DB_AFTER</a>, <a href="../api_reference/C/dbcput.html#put_DB_BEFORE" class="olink">DB_BEFORE</a>, <a href="../api_reference/C/dbcput.html#put_DB_KEYFIRST" class="olink">DB_KEYFIRST</a> or <a href="../api_reference/C/dbcput.html#put_DB_KEYLAST" class="olink">DB_KEYLAST</a> flags.
Alternatively, Berkeley DB may be configured to sort duplicate data items.</p>
<p>When stepping through the database sequentially, duplicate data items will
be returned individually, as a key/data pair, where the key item only
changes after the last duplicate data item has been returned. For this
reason, duplicate data items cannot be accessed using the
<a href="../api_reference/C/dbget.html" class="olink">DB->get()</a> method, as it always returns the first of the duplicate data
items. Duplicate data items should be retrieved using a Berkeley DB cursor
interface such as the <a href="../api_reference/C/dbcget.html" class="olink">DBC->get()</a> method.</p>
<p>There is a flag that permits applications to request the following data
item only if it <span class="bold"><strong>is</strong></span> a duplicate data item of the current entry,
see <a href="../api_reference/C/dbcget.html#dbcget_DB_NEXT_DUP" class="olink">DB_NEXT_DUP</a> for more information. There is a flag that
permits applications to request the following data item only if it
<span class="bold"><strong>is not</strong></span> a duplicate data item of the current entry, see
<a href="../api_reference/C/dbcget.html#dbcget_DB_NEXT_NODUP" class="olink">DB_NEXT_NODUP</a> and <a href="../api_reference/C/dbcget.html#dbcget_DB_PREV_NODUP" class="olink">DB_PREV_NODUP</a> for more information.</p>
<p>It is also possible to maintain duplicate records in sorted order. Sorting
duplicates will significantly increase performance when searching them
and performing equality joins, common operations when using secondary
indices. To configure Berkeley DB to sort duplicate data items, the application
must call the <a href="../api_reference/C/dbset_flags.html" class="olink">DB->set_flags()</a> method with the <a href="../api_reference/C/dbset_flags.html#dbset_flags_DB_DUPSORT" class="olink">DB_DUPSORT</a> flag (in
addition to the <a href="../api_reference/C/dbset_flags.html#dbset_flags_DB_DUP" class="olink">DB_DUP</a> flag). In addition, a custom comparison
function may be specified using the <a href="../api_reference/C/dbset_dup_compare.html" class="olink">DB->set_dup_compare()</a> method. If the
<a href="../api_reference/C/dbset_flags.html#dbset_flags_DB_DUPSORT" class="olink">DB_DUPSORT</a> flag is given, but no comparison routine is specified,
then Berkeley DB defaults to the same lexicographical sorting used for Btree
keys, with shorter items collating before longer items.</p>
<p>If the duplicate data items are unsorted, applications may store identical
duplicate data items, or, for those that just like the way it sounds,
<span class="emphasis"><em>duplicate duplicates</em></span>.</p>
<p><span class="bold"><strong>In this release it is an error to attempt to store identical
duplicate data items when duplicates are being stored in a sorted order.</strong></span>
This restriction is expected to be lifted in a future release. There
is a flag that permits applications to disallow storing duplicate data
items when the database has been configured for sorted duplicates, see
<a href="../api_reference/C/dbput.html#put_DB_NODUPDATA" class="olink">DB_NODUPDATA</a> for more information. Applications not wanting to
permit duplicate duplicates in databases configured for sorted
duplicates should begin using the <a href="../api_reference/C/dbput.html#put_DB_NODUPDATA" class="olink">DB_NODUPDATA</a> flag immediately.</p>
<p>For further information on how searching and insertion behaves in the
presence of duplicates (sorted or not), see the <a href="../api_reference/C/dbget.html" class="olink">DB->get()</a> <a href="../api_reference/C/dbput.html" class="olink">DB->put()</a>, <a href="../api_reference/C/dbcget.html" class="olink">DBC->get()</a> and
<a href="../api_reference/C/dbcput.html" class="olink">DBC->put()</a> documentation.</p>
</div>
<div class="sect2" lang="en" xml:lang="en">
<div class="titlepage">
<div>
<div>
<h3 class="title"><a id="am_conf_malloc"></a>Non-local memory allocation</h3>
</div>
</div>
</div>
<p>Berkeley DB allocates memory for returning key/data pairs and statistical
information which becomes the responsibility of the application.
There are also interfaces where an application will allocate memory
which becomes the responsibility of Berkeley DB.</p>
<p>On systems in which there may be multiple library versions of the
standard allocation routines (notably Windows NT), transferring memory
between the library and the application will fail because the Berkeley DB
library allocates memory from a different heap than the application
uses to free it, or vice versa. To avoid this problem, the
<a href="../api_reference/C/envset_alloc.html" class="olink">DB_ENV->set_alloc()</a> and <a href="../api_reference/C/dbset_alloc.html" class="olink">DB->set_alloc()</a> methods can be used to
give Berkeley DB references to the application's allocation routines.</p>
</div>
</div>
<div class="navfooter">
<hr />
<table width="100%" summary="Navigation footer">
<tr>
<td width="40%" align="left"><a accesskey="p" href="am_conf_logrec.html">Prev</a> </td>
<td width="20%" align="center">
<a accesskey="u" href="am_conf.html">Up</a>
</td>
<td width="40%" align="right"> <a accesskey="n" href="bt_conf.html">Next</a></td>
</tr>
<tr>
<td width="40%" align="left" valign="top">Logical record numbers </td>
<td width="20%" align="center">
<a accesskey="h" href="index.html">Home</a>
</td>
<td width="40%" align="right" valign="top"> Btree access method specific configuration</td>
</tr>
</table>
</div>
</body>
</html>
|