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
|
<! "@(#)intro.so 10.2 (Sleepycat) 12/18/98">
<!Copyright 1997, 1998 by Sleepycat Software, Inc. All rights reserved.>
<html>
<body bgcolor=white>
<head>
<title>Berkeley DB Reference Guide: Memory Pool Subsystem</title>
<meta name="description" content="Berkeley DB: An embedded database programmatic toolkit.">
<meta name="keywords" content="embedded,database,programmatic,toolkit,b+tree,btr
ee,hash,hashing,transaction,transactions,locking,logging,access method,access me
thods,java,C,C++">
</head>
<h3>Berkeley DB Reference Guide: Memory Pool Subsystem</h3>
<p>
<h1 align=center>Berkeley DB and the memory pool</h1>
<p>
The memory pool subsystem is the general-purpose shared memory buffer pool
used by Berkeley DB. This module is useful outside of the Berkeley DB package for
processes that require page-oriented, cached, shared file access.
<p>
A memory pool is a shared memory cache shared by any number of processes
and threads within processes. The <a href="../../api_c/DbMpool/open.html">memp_open</a> interface opens, and
optionally creates, a memory pool. When that pool is no longer in use,
it should be closed, using the <a href="../../api_c/DbMpool/close.html">memp_close</a> interface.
<p>
The <a href="../../api_c/DbMpoolFile/open.html">memp_fopen</a> interface opens an underlying file within the
memory pool. When that file is no longer in use, it should be closed,
using the <a href="../../api_c/DbMpoolFile/close.html">memp_fclose</a> interface. The <a href="../../api_c/DbMpoolFile/get.html">memp_fget</a> interface
is used to retrieve pages from files in the pool. All retrieved pages
must be subsequently returned using the <a href="../../api_c/DbMpoolFile/put.html">memp_fput</a> interface. At
the time that pages are returned, they may be marked <b>dirty</b>, which
causes them to be written to the backing disk file before being discarded
from the pool. If there is insufficient room to bring a new page in the
pool, a page is selected to be discarded from the pool. If that page is
dirty, it is first written to the backing file. The page is selected
using a somewhat modified least-recently-used algorithm. Pages in files
may also be explicitly marked clean or dirty using the <a href="../../api_c/DbMpoolFile/set.html">memp_fset</a>
interface. All dirty pages in the pool from any underlying file may also
be flushed as a group using the <a href="../../api_c/DbMpoolFile/sync.html">memp_fsync</a> interface.
<p>
There are additional interfaces for manipulating the entire memory pool:
<p>
It is possible to gradually flush buffers from the pool in order to
maintain a consistent percentage of clean buffers in the pool using the
<a href="../../api_c/DbMpool/trickle.html">memp_trickle</a> interface.
<p>
The <a href="../../utility/db_stat.html">db_stat</a> utility uses the <a href="../../api_c/DbMpool/stat.html">memp_stat</a> interface to
display statistics about the efficiency of the pool.
<p>
As some conversion may be necessary when pages are read into, or written
from, the pool, the <a href="../../api_c/DbMpool/db_register.html">memp_register</a> function allows applications to
specify automatic input and output processing in these cases.
<p>
There is one additional interface that is intended for manipulating the
memory pool, but which is specific to database systems. The
<a href="../../api_c/DbMpool/sync.html">memp_sync</a> interface flushes dirty pages from all files held in
the pool up to the current database log sequence number.
<p>
Finally, the entire pool may be discarded using the <a href="../../api_c/DbMpool/unlink.html">memp_unlink</a>
interface.
<p>
<a href="../../ref/txn/other.html"><img src="../../images/prev.gif"></a>
<a href="../../ref/toc.html"><img src="../../images/toc.gif"></a>
<a href="../../ref/mp/config.html"><img src="../../images/next.gif"></a>
</tt>
</body>
</html>
|
