mirror of
https://github.com/berkeleydb/libdb.git
synced 2024-11-16 17:16:25 +00:00
160 lines
6.3 KiB
HTML
160 lines
6.3 KiB
HTML
<?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>Reverse BTree Splits</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="Getting Started with Berkeley DB Transaction Processing" />
|
||
<link rel="up" href="txnconcurrency.html" title="Chapter 4. Concurrency" />
|
||
<link rel="prev" href="txnnowait.html" title="No Wait on Blocks" />
|
||
<link rel="next" href="filemanagement.html" title="Chapter 5. Managing DB Files" />
|
||
</head>
|
||
<body>
|
||
<div xmlns="" class="navheader">
|
||
<div class="libver">
|
||
<p>Library Version 11.2.5.3</p>
|
||
</div>
|
||
<table width="100%" summary="Navigation header">
|
||
<tr>
|
||
<th colspan="3" align="center">Reverse BTree Splits</th>
|
||
</tr>
|
||
<tr>
|
||
<td width="20%" align="left"><a accesskey="p" href="txnnowait.html">Prev</a> </td>
|
||
<th width="60%" align="center">Chapter 4. Concurrency</th>
|
||
<td width="20%" align="right"> <a accesskey="n" href="filemanagement.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="reversesplit"></a>Reverse BTree Splits</h2>
|
||
</div>
|
||
</div>
|
||
</div>
|
||
<p>
|
||
If your application is using the Btree access method, and your
|
||
application is repeatedly deleting then adding records to your
|
||
database, then you might be able to reduce lock contention by
|
||
turning off reverse Btree splits.
|
||
</p>
|
||
<p>
|
||
As pages are emptied in a database, DB attempts to
|
||
delete empty pages in order to keep
|
||
the database as small as possible and minimize search time.
|
||
Moreover, when a page in the database fills
|
||
up, DB, of course, adds additional pages
|
||
to make room for more data.
|
||
</p>
|
||
<p>
|
||
Adding and deleting pages in the database requires that the
|
||
writing thread lock the parent page. Consequently, as the
|
||
number of pages in your database diminishes, your application
|
||
will see increasingly more lock contention; the maximum level
|
||
of concurrency in a database of two pages is far smaller than
|
||
that in a database of 100 pages, because there are fewer pages
|
||
that can be locked.
|
||
</p>
|
||
<p>
|
||
Therefore, if you prevent the database from being reduced to a
|
||
minimum number of pages, you can improve your application's
|
||
concurrency throughput. Note, however, that you should do so
|
||
only if your application tends to delete and then add the same
|
||
data. If this is not the case, then preventing reverse Btree
|
||
splits can harm your database search time.
|
||
</p>
|
||
<p>
|
||
To turn off reverse Btree splits,
|
||
<span>
|
||
provide the
|
||
<code class="literal">DB_REVSPLITOFF</code> flag to the
|
||
|
||
<code class="methodname">Db::set_flags()</code>
|
||
method.
|
||
</span>
|
||
|
||
</p>
|
||
<p>
|
||
For example:
|
||
</p>
|
||
<pre class="programlisting">#include "db_cxx.h"
|
||
|
||
...
|
||
|
||
int main(void)
|
||
{
|
||
u_int32_t env_flags = DB_CREATE | // If the environment does not
|
||
// exist, create it.
|
||
DB_INIT_LOCK | // Initialize locking
|
||
DB_INIT_LOG | // Initialize locking
|
||
DB_INIT_MPOOL | // Initialize the cache
|
||
DB_THREAD | // Free-thread the env handle
|
||
DB_INIT_TXN; // Initialize transactions
|
||
|
||
u_int32_t db_flags = DB_CREATE | DB_AUTO_COMMIT;
|
||
Db *dbp = NULL;
|
||
const char *file_name = "mydb.db";
|
||
|
||
std::string envHome("/export1/testEnv");
|
||
DbEnv myEnv(0);
|
||
|
||
try {
|
||
|
||
myEnv.open(envHome.c_str(), env_flags, 0);
|
||
dbp = new Db(&myEnv, 0);
|
||
|
||
// Turn off BTree reverse split.
|
||
dbp=>set_flags(DB_REVSPLITOFF);
|
||
|
||
dbp->open(dbp, // Pointer to the database
|
||
NULL, // Txn pointer
|
||
file_name, // File name
|
||
NULL, // Logical db name
|
||
DB_BTREE, // Database type (using btree)
|
||
db_flags, // Open flags
|
||
0); // File mode. Using defaults
|
||
|
||
} catch(DbException &e) {
|
||
std::cerr << "Error opening database and environment: "
|
||
<< file_name << ", " << envHome << std::endl;
|
||
std::cerr << e.what() << std::endl;
|
||
}
|
||
|
||
try {
|
||
dbp->close(dbp, 0);
|
||
myEnv.close(0);
|
||
} catch(DbException &e) {
|
||
std::cerr << "Error closing database and environment: "
|
||
<< file_name << ", " << envHome << std::endl;
|
||
std::cerr << e.what() << std::endl;
|
||
return (EXIT_FAILURE);
|
||
}
|
||
|
||
return (EXIT_SUCCESS);
|
||
} </pre>
|
||
</div>
|
||
<div class="navfooter">
|
||
<hr />
|
||
<table width="100%" summary="Navigation footer">
|
||
<tr>
|
||
<td width="40%" align="left"><a accesskey="p" href="txnnowait.html">Prev</a> </td>
|
||
<td width="20%" align="center">
|
||
<a accesskey="u" href="txnconcurrency.html">Up</a>
|
||
</td>
|
||
<td width="40%" align="right"> <a accesskey="n" href="filemanagement.html">Next</a></td>
|
||
</tr>
|
||
<tr>
|
||
<td width="40%" align="left" valign="top">No Wait on Blocks </td>
|
||
<td width="20%" align="center">
|
||
<a accesskey="h" href="index.html">Home</a>
|
||
</td>
|
||
<td width="40%" align="right" valign="top"> Chapter 5. Managing DB Files</td>
|
||
</tr>
|
||
</table>
|
||
</div>
|
||
</body>
|
||
</html>
|