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
|
<?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>59.5. Row Locking in Foreign Data Wrappers</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 Vsnapshot" /><link rel="prev" href="fdw-planning.html" title="59.4. Foreign Data Wrapper Query Planning" /><link rel="next" href="tablesample-method.html" title="Chapter 60. Writing a Table Sampling Method" /></head><body id="docContent" class="container-fluid col-10"><div class="navheader"><table width="100%" summary="Navigation header"><tr><th colspan="5" align="center">59.5. Row Locking in Foreign Data Wrappers</th></tr><tr><td width="10%" align="left"><a accesskey="p" href="fdw-planning.html" title="59.4. Foreign Data Wrapper Query Planning">Prev</a> </td><td width="10%" align="left"><a accesskey="u" href="fdwhandler.html" title="Chapter 59. Writing a Foreign Data Wrapper">Up</a></td><th width="60%" align="center">Chapter 59. Writing a Foreign Data Wrapper</th><td width="10%" align="right"><a accesskey="h" href="index.html" title="PostgreSQL 15.5 Documentation">Home</a></td><td width="10%" align="right"> <a accesskey="n" href="tablesample-method.html" title="Chapter 60. Writing a Table Sampling Method">Next</a></td></tr></table><hr /></div><div class="sect1" id="FDW-ROW-LOCKING"><div class="titlepage"><div><div><h2 class="title" style="clear: both">59.5. Row Locking in Foreign Data Wrappers</h2></div></div></div><p>
If an FDW's underlying storage mechanism has a concept of locking
individual rows to prevent concurrent updates of those rows, it is
usually worthwhile for the FDW to perform row-level locking with as
close an approximation as practical to the semantics used in
ordinary <span class="productname">PostgreSQL</span> tables. There are multiple
considerations involved in this.
</p><p>
One key decision to be made is whether to perform <em class="firstterm">early
locking</em> or <em class="firstterm">late locking</em>. In early locking, a row is
locked when it is first retrieved from the underlying store, while in
late locking, the row is locked only when it is known that it needs to
be locked. (The difference arises because some rows may be discarded by
locally-checked restriction or join conditions.) Early locking is much
simpler and avoids extra round trips to a remote store, but it can cause
locking of rows that need not have been locked, resulting in reduced
concurrency or even unexpected deadlocks. Also, late locking is only
possible if the row to be locked can be uniquely re-identified later.
Preferably the row identifier should identify a specific version of the
row, as <span class="productname">PostgreSQL</span> TIDs do.
</p><p>
By default, <span class="productname">PostgreSQL</span> ignores locking considerations
when interfacing to FDWs, but an FDW can perform early locking without
any explicit support from the core code. The API functions described
in <a class="xref" href="fdw-callbacks.html#FDW-CALLBACKS-ROW-LOCKING" title="59.2.6. FDW Routines for Row Locking">Section 59.2.6</a>, which were added
in <span class="productname">PostgreSQL</span> 9.5, allow an FDW to use late locking if
it wishes.
</p><p>
An additional consideration is that in <code class="literal">READ COMMITTED</code>
isolation mode, <span class="productname">PostgreSQL</span> may need to re-check
restriction and join conditions against an updated version of some
target tuple. Rechecking join conditions requires re-obtaining copies
of the non-target rows that were previously joined to the target tuple.
When working with standard <span class="productname">PostgreSQL</span> tables, this is
done by including the TIDs of the non-target tables in the column list
projected through the join, and then re-fetching non-target rows when
required. This approach keeps the join data set compact, but it
requires inexpensive re-fetch capability, as well as a TID that can
uniquely identify the row version to be re-fetched. By default,
therefore, the approach used with foreign tables is to include a copy of
the entire row fetched from a foreign table in the column list projected
through the join. This puts no special demands on the FDW but can
result in reduced performance of merge and hash joins. An FDW that is
capable of meeting the re-fetch requirements can choose to do it the
first way.
</p><p>
For an <code class="command">UPDATE</code> or <code class="command">DELETE</code> on a foreign table, it
is recommended that the <code class="literal">ForeignScan</code> operation on the target
table perform early locking on the rows that it fetches, perhaps via the
equivalent of <code class="command">SELECT FOR UPDATE</code>. An FDW can detect whether
a table is an <code class="command">UPDATE</code>/<code class="command">DELETE</code> target at plan time
by comparing its relid to <code class="literal">root->parse->resultRelation</code>,
or at execution time by using <code class="function">ExecRelationIsTargetRelation()</code>.
An alternative possibility is to perform late locking within the
<code class="function">ExecForeignUpdate</code> or <code class="function">ExecForeignDelete</code>
callback, but no special support is provided for this.
</p><p>
For foreign tables that are specified to be locked by a <code class="command">SELECT
FOR UPDATE/SHARE</code> command, the <code class="literal">ForeignScan</code> operation can
again perform early locking by fetching tuples with the equivalent
of <code class="command">SELECT FOR UPDATE/SHARE</code>. To perform late locking
instead, provide the callback functions defined
in <a class="xref" href="fdw-callbacks.html#FDW-CALLBACKS-ROW-LOCKING" title="59.2.6. FDW Routines for Row Locking">Section 59.2.6</a>.
In <code class="function">GetForeignRowMarkType</code>, select rowmark option
<code class="literal">ROW_MARK_EXCLUSIVE</code>, <code class="literal">ROW_MARK_NOKEYEXCLUSIVE</code>,
<code class="literal">ROW_MARK_SHARE</code>, or <code class="literal">ROW_MARK_KEYSHARE</code> depending
on the requested lock strength. (The core code will act the same
regardless of which of these four options you choose.)
Elsewhere, you can detect whether a foreign table was specified to be
locked by this type of command by using <code class="function">get_plan_rowmark</code> at
plan time, or <code class="function">ExecFindRowMark</code> at execution time; you must
check not only whether a non-null rowmark struct is returned, but that
its <code class="structfield">strength</code> field is not <code class="literal">LCS_NONE</code>.
</p><p>
Lastly, for foreign tables that are used in an <code class="command">UPDATE</code>,
<code class="command">DELETE</code> or <code class="command">SELECT FOR UPDATE/SHARE</code> command but
are not specified to be row-locked, you can override the default choice
to copy entire rows by having <code class="function">GetForeignRowMarkType</code> select
option <code class="literal">ROW_MARK_REFERENCE</code> when it sees lock strength
<code class="literal">LCS_NONE</code>. This will cause <code class="function">RefetchForeignRow</code> to
be called with that value for <code class="structfield">markType</code>; it should then
re-fetch the row without acquiring any new lock. (If you have
a <code class="function">GetForeignRowMarkType</code> function but don't wish to re-fetch
unlocked rows, select option <code class="literal">ROW_MARK_COPY</code>
for <code class="literal">LCS_NONE</code>.)
</p><p>
See <code class="filename">src/include/nodes/lockoptions.h</code>, the comments
for <code class="type">RowMarkType</code> and <code class="type">PlanRowMark</code>
in <code class="filename">src/include/nodes/plannodes.h</code>, and the comments for
<code class="type">ExecRowMark</code> in <code class="filename">src/include/nodes/execnodes.h</code> for
additional information.
</p></div><div class="navfooter"><hr /><table width="100%" summary="Navigation footer"><tr><td width="40%" align="left"><a accesskey="p" href="fdw-planning.html" title="59.4. Foreign Data Wrapper Query Planning">Prev</a> </td><td width="20%" align="center"><a accesskey="u" href="fdwhandler.html" title="Chapter 59. Writing a Foreign Data Wrapper">Up</a></td><td width="40%" align="right"> <a accesskey="n" href="tablesample-method.html" title="Chapter 60. Writing a Table Sampling Method">Next</a></td></tr><tr><td width="40%" align="left" valign="top">59.4. Foreign Data Wrapper Query Planning </td><td width="20%" align="center"><a accesskey="h" href="index.html" title="PostgreSQL 15.5 Documentation">Home</a></td><td width="40%" align="right" valign="top"> Chapter 60. Writing a Table Sampling Method</td></tr></table></div></body></html>
|
