Baoding-FenShuaJiang
/
Green_GoodERP18_FSJ_Standard


			
				
					
						
						
							
							<?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>68.6. Database Page Layout</title><link rel="stylesheet" type="text/css" href="stylesheet.css" /><link rev="made" href="pgsql-docs@lists.postgresql.org" /><meta name="generator" content="DocBook XSL Stylesheets V1.79.1" /><link rel="prev" href="storage-init.html" title="68.5. The Initialization Fork" /><link rel="next" href="bki.html" title="Chapter 69. System Catalog Declarations and Initial Contents" /></head><body><div xmlns="http://www.w3.org/TR/xhtml1/transitional" class="navheader"><table width="100%" summary="Navigation header"><tr><th colspan="5" align="center">68.6. Database Page Layout</th></tr><tr><td width="10%" align="left"><a accesskey="p" href="storage-init.html" title="68.5. The Initialization Fork">Prev</a> </td><td width="10%" align="left"><a accesskey="u" href="storage.html" title="Chapter 68. Database Physical Storage">Up</a></td><th width="60%" align="center">Chapter 68. Database Physical Storage</th><td width="10%" align="right"><a accesskey="h" href="index.html" title="PostgreSQL 12.4 Documentation">Home</a></td><td width="10%" align="right"> <a accesskey="n" href="bki.html" title="Chapter 69. System Catalog Declarations and Initial Contents">Next</a></td></tr></table><hr></hr></div><div class="sect1" id="STORAGE-PAGE-LAYOUT"><div class="titlepage"><div><div><h2 class="title" style="clear: both">68.6. Database Page Layout</h2></div></div></div><div class="toc"><dl class="toc"><dt><span class="sect2"><a href="storage-page-layout.html#STORAGE-TUPLE-LAYOUT">68.6.1. Table Row Layout</a></span></dt></dl></div><p>
This section provides an overview of the page format used within
<span class="productname">PostgreSQL</span> tables and indexes.<a href="#ftn.id-1.10.21.8.2.2" class="footnote"><sup class="footnote" id="id-1.10.21.8.2.2">[15]</sup></a>
Sequences and <acronym class="acronym">TOAST</acronym> tables are formatted just like a regular table.
</p><p>
In the following explanation, a
<em class="firstterm">byte</em>
is assumed to contain 8 bits.  In addition, the term
<em class="firstterm">item</em>
refers to an individual data value that is stored on a page.  In a table,
an item is a row; in an index, an item is an index entry.
</p><p>
Every table and index is stored as an array of <em class="firstterm">pages</em> of a
fixed size (usually 8 kB, although a different page size can be selected
when compiling the server).  In a table, all the pages are logically
equivalent, so a particular item (row) can be stored in any page.  In
indexes, the first page is generally reserved as a <em class="firstterm">metapage</em>
holding control information, and there can be different types of pages
within the index, depending on the index access method.
</p><p>
<a class="xref" href="storage-page-layout.html#PAGE-TABLE" title="Table 68.2. Overall Page Layout">Table 68.2</a> shows the overall layout of a page.
There are five parts to each page.
</p><div class="table" id="PAGE-TABLE"><p class="title"><strong>Table 68.2. Overall Page Layout</strong></p><div class="table-contents"><table class="table" summary="Overall Page Layout" border="1"><colgroup><col /><col /></colgroup><thead><tr><th>
Item
</th><th>Description</th></tr></thead><tbody><tr><td>PageHeaderData</td><td>24 bytes long. Contains general information about the page, including
free space pointers.</td></tr><tr><td>ItemIdData</td><td>Array of item identifiers pointing to the actual items. Each
entry is an (offset,length) pair. 4 bytes per item.</td></tr><tr><td>Free space</td><td>The unallocated space. New item identifiers are allocated from
the start of this area, new items from the end.</td></tr><tr><td>Items</td><td>The actual items themselves.</td></tr><tr><td>Special space</td><td>Index access method specific data. Different methods store different
data. Empty in ordinary tables.</td></tr></tbody></table></div></div><br class="table-break" /><p>

  The first 24 bytes of each page consists of a page header
  (<code class="structname">PageHeaderData</code>). Its format is detailed in <a class="xref" href="storage-page-layout.html#PAGEHEADERDATA-TABLE" title="Table 68.3. PageHeaderData Layout">Table 68.3</a>. The first field tracks the most
  recent WAL entry related to this page. The second field contains
  the page checksum if <a class="xref" href="app-initdb.html#APP-INITDB-DATA-CHECKSUMS">data checksums</a> are
  enabled.  Next is a 2-byte field containing flag bits. This is followed
  by three 2-byte integer fields (<code class="structfield">pd_lower</code>,
  <code class="structfield">pd_upper</code>, and
  <code class="structfield">pd_special</code>).  These contain byte offsets
  from the page start to the start of unallocated space, to the end of
  unallocated space, and to the start of the special space.  The next 2
  bytes of the page header, <code class="structfield">pd_pagesize_version</code>,
  store both the page size and a version indicator.  Beginning with
  <span class="productname">PostgreSQL</span> 8.3 the version number is 4;
  <span class="productname">PostgreSQL</span> 8.1 and 8.2 used version number 3;
  <span class="productname">PostgreSQL</span> 8.0 used version number 2;
  <span class="productname">PostgreSQL</span> 7.3 and 7.4 used version number 1;
  prior releases used version number 0.
  (The basic page layout and header format has not changed in most of these
  versions, but the layout of heap row headers has.)  The page size
  is basically only present as a cross-check; there is no support for having
  more than one page size in an installation.
  The last field is a hint that shows whether pruning the page is likely
  to be profitable: it tracks the oldest un-pruned XMAX on the page.

 </p><div class="table" id="PAGEHEADERDATA-TABLE"><p class="title"><strong>Table 68.3. PageHeaderData Layout</strong></p><div class="table-contents"><table class="table" summary="PageHeaderData Layout" border="1"><colgroup><col /><col /><col /><col /></colgroup><thead><tr><th>Field</th><th>Type</th><th>Length</th><th>Description</th></tr></thead><tbody><tr><td>pd_lsn</td><td>PageXLogRecPtr</td><td>8 bytes</td><td>LSN: next byte after last byte of WAL record for last change
   to this page</td></tr><tr><td>pd_checksum</td><td>uint16</td><td>2 bytes</td><td>Page checksum</td></tr><tr><td>pd_flags</td><td>uint16</td><td>2 bytes</td><td>Flag bits</td></tr><tr><td>pd_lower</td><td>LocationIndex</td><td>2 bytes</td><td>Offset to start of free space</td></tr><tr><td>pd_upper</td><td>LocationIndex</td><td>2 bytes</td><td>Offset to end of free space</td></tr><tr><td>pd_special</td><td>LocationIndex</td><td>2 bytes</td><td>Offset to start of special space</td></tr><tr><td>pd_pagesize_version</td><td>uint16</td><td>2 bytes</td><td>Page size and layout version number information</td></tr><tr><td>pd_prune_xid</td><td>TransactionId</td><td>4 bytes</td><td>Oldest unpruned XMAX on page, or zero if none</td></tr></tbody></table></div></div><br class="table-break" /><p>
  All the details can be found in
  <code class="filename">src/include/storage/bufpage.h</code>.
 </p><p>
  Following the page header are item identifiers
  (<code class="type">ItemIdData</code>), each requiring four bytes.
  An item identifier contains a byte-offset to
  the start of an item, its length in bytes, and a few attribute bits
  which affect its interpretation.
  New item identifiers are allocated
  as needed from the beginning of the unallocated space.
  The number of item identifiers present can be determined by looking at
  <code class="structfield">pd_lower</code>, which is increased to allocate a new identifier.
  Because an item
  identifier is never moved until it is freed, its index can be used on a
  long-term basis to reference an item, even when the item itself is moved
  around on the page to compact free space.  In fact, every pointer to an
  item (<code class="type">ItemPointer</code>, also known as
  <code class="type">CTID</code>) created by
  <span class="productname">PostgreSQL</span> consists of a page number and the
  index of an item identifier.

 </p><p>

  The items themselves are stored in space allocated backwards from the end
  of unallocated space.  The exact structure varies depending on what the
  table is to contain. Tables and sequences both use a structure named
  <code class="type">HeapTupleHeaderData</code>, described below.

 </p><p>

  The final section is the <span class="quote">“<span class="quote">special section</span>”</span> which can
  contain anything the access method wishes to store.  For example,
  b-tree indexes store links to the page's left and right siblings,
  as well as some other data relevant to the index structure.
  Ordinary tables do not use a special section at all (indicated by setting
  <code class="structfield">pd_special</code> to equal the page size).

 </p><p>
  <a class="xref" href="storage-page-layout.html#STORAGE-PAGE-LAYOUT-FIGURE" title="Figure 68.1. Page Layout">Figure 68.1</a> illustrates how these parts are
  laid out in a page.
 </p><div class="figure" id="STORAGE-PAGE-LAYOUT-FIGURE"><p class="title"><strong>Figure 68.1. Page Layout</strong></p><div class="figure-contents"><div class="mediaobject"><object type="image/svg+xml" data="pagelayout.svg" width="100%"></object></div></div></div><br class="figure-break" /><div class="sect2" id="STORAGE-TUPLE-LAYOUT"><div class="titlepage"><div><div><h3 class="title">68.6.1. Table Row Layout</h3></div></div></div><p>

  All table rows are structured in the same way. There is a fixed-size
  header (occupying 23 bytes on most machines), followed by an optional null
  bitmap, an optional object ID field, and the user data. The header is
  detailed
  in <a class="xref" href="storage-page-layout.html#HEAPTUPLEHEADERDATA-TABLE" title="Table 68.4. HeapTupleHeaderData Layout">Table 68.4</a>.  The actual user data
  (columns of the row) begins at the offset indicated by
  <code class="structfield">t_hoff</code>, which must always be a multiple of the MAXALIGN
  distance for the platform.
  The null bitmap is
  only present if the <em class="firstterm">HEAP_HASNULL</em> bit is set in
  <code class="structfield">t_infomask</code>. If it is present it begins just after
  the fixed header and occupies enough bytes to have one bit per data column
  (that is, the number of bits that equals the attribute count in
  <code class="structfield">t_infomask2</code>). In this list of bits, a
  1 bit indicates not-null, a 0 bit is a null.  When the bitmap is not
  present, all columns are assumed not-null.
  The object ID is only present if the <em class="firstterm">HEAP_HASOID_OLD</em> bit
  is set in <code class="structfield">t_infomask</code>.  If present, it appears just
  before the <code class="structfield">t_hoff</code> boundary.  Any padding needed to make
  <code class="structfield">t_hoff</code> a MAXALIGN multiple will appear between the null
  bitmap and the object ID.  (This in turn ensures that the object ID is
  suitably aligned.)

 </p><div class="table" id="HEAPTUPLEHEADERDATA-TABLE"><p class="title"><strong>Table 68.4. HeapTupleHeaderData Layout</strong></p><div class="table-contents"><table class="table" summary="HeapTupleHeaderData Layout" border="1"><colgroup><col /><col /><col /><col /></colgroup><thead><tr><th>Field</th><th>Type</th><th>Length</th><th>Description</th></tr></thead><tbody><tr><td>t_xmin</td><td>TransactionId</td><td>4 bytes</td><td>insert XID stamp</td></tr><tr><td>t_xmax</td><td>TransactionId</td><td>4 bytes</td><td>delete XID stamp</td></tr><tr><td>t_cid</td><td>CommandId</td><td>4 bytes</td><td>insert and/or delete CID stamp (overlays with t_xvac)</td></tr><tr><td>t_xvac</td><td>TransactionId</td><td>4 bytes</td><td>XID for VACUUM operation moving a row version</td></tr><tr><td>t_ctid</td><td>ItemPointerData</td><td>6 bytes</td><td>current TID of this or newer row version</td></tr><tr><td>t_infomask2</td><td>uint16</td><td>2 bytes</td><td>number of attributes, plus various flag bits</td></tr><tr><td>t_infomask</td><td>uint16</td><td>2 bytes</td><td>various flag bits</td></tr><tr><td>t_hoff</td><td>uint8</td><td>1 byte</td><td>offset to user data</td></tr></tbody></table></div></div><br class="table-break" /><p>
   All the details can be found in
   <code class="filename">src/include/access/htup_details.h</code>.
 </p><p>

  Interpreting the actual data can only be done with information obtained
  from other tables, mostly <code class="structname">pg_attribute</code>. The
  key values needed to identify field locations are
  <code class="structfield">attlen</code> and <code class="structfield">attalign</code>.
  There is no way to directly get a
  particular attribute, except when there are only fixed width fields and no
  null values. All this trickery is wrapped up in the functions
  <em class="firstterm">heap_getattr</em>, <em class="firstterm">fastgetattr</em>
  and <em class="firstterm">heap_getsysattr</em>.

 </p><p>

  To read the data you need to examine each attribute in turn. First check
  whether the field is NULL according to the null bitmap. If it is, go to
  the next. Then make sure you have the right alignment.  If the field is a
  fixed width field, then all the bytes are simply placed. If it's a
  variable length field (attlen = -1) then it's a bit more complicated.
  All variable-length data types share the common header structure
  <code class="type">struct varlena</code>, which includes the total length of the stored
  value and some flag bits.  Depending on the flags, the data can be either
  inline or in a <acronym class="acronym">TOAST</acronym> table;
  it might be compressed, too (see <a class="xref" href="storage-toast.html" title="68.2. TOAST">Section 68.2</a>).

 </p></div><div class="footnotes"><br /><hr style="width:100; text-align:left;margin-left: 0" /><div id="ftn.id-1.10.21.8.2.2" class="footnote"><p><a href="#id-1.10.21.8.2.2" class="para"><sup class="para">[15] </sup></a>
    Actually, use of this page format is not required for either table or
    index access methods. The <code class="literal">heap</code> table access method
    always uses this format.  All the existing index methods also use the
    basic format, but the data kept on index metapages usually doesn't follow
    the item layout rules.
  </p></div></div></div><div class="navfooter"><hr /><table width="100%" summary="Navigation footer"><tr><td width="40%" align="left"><a accesskey="p" href="storage-init.html">Prev</a> </td><td width="20%" align="center"><a accesskey="u" href="storage.html">Up</a></td><td width="40%" align="right"> <a accesskey="n" href="bki.html">Next</a></td></tr><tr><td width="40%" align="left" valign="top">68.5. The Initialization Fork </td><td width="20%" align="center"><a accesskey="h" href="index.html">Home</a></td><td width="40%" align="right" valign="top"> Chapter 69. System Catalog Declarations and Initial Contents</td></tr></table></div></body></html>