Fix multiple bugs in contrib/pgstattuple's pgstatindex() function.

Dead or half-dead index leaf pages were incorrectly reported as live, as a
consequence of a code rearrangement I made (during a moment of severe brain
fade, evidently) in commit d287818eb5.

The index metapage was not counted in index_size, causing that result to
not agree with the actual index size on-disk.

Index root pages were not counted in internal_pages, which is inconsistent
compared to the case of a root that's also a leaf (one-page index), where
the root would be counted in leaf_pages.  Aside from that inconsistency,
this could lead to additional transient discrepancies between the reported
page counts and index_size, since it's possible for pgstatindex's scan to
see zero or multiple pages marked as BTP_ROOT, if the root moves due to
a split during the scan.  With these fixes, index_size will always be
exactly one page more than the sum of the displayed page counts.

Also, the index_size result was incorrectly documented as being measured in
pages; it's always been measured in bytes.  (While fixing that, I couldn't
resist doing some small additional wordsmithing on the pgstattuple docs.)

Including the metapage causes the reported index_size to not be zero for
an empty index.  To preserve the desired property that the pgstattuple
regression test results are platform-independent (ie, BLCKSZ configuration
independent), scale the index_size result in the regression tests.

The documentation issue was reported by Otsuka Kenji, and the inconsistent
root page counting by Peter Geoghegan; the other problems noted by me.
Back-patch to all supported branches, because this has been broken for
a long time.
This commit is contained in:
Tom Lane 2016-02-18 15:40:35 -05:00
parent 18777c38e9
commit 48e6c943e5
4 changed files with 68 additions and 47 deletions

View File

@ -41,40 +41,44 @@ select pgstattuple(relname) from pg_class where relname = 'test';
(0,0,0,0,0,0,0,0,0)
(1 row)
select * from pgstatindex('test_pkey');
select version, tree_level,
index_size / current_setting('block_size')::int as index_size,
root_block_no, internal_pages, leaf_pages, empty_pages, deleted_pages,
avg_leaf_density, leaf_fragmentation
from pgstatindex('test_pkey');
version | tree_level | index_size | root_block_no | internal_pages | leaf_pages | empty_pages | deleted_pages | avg_leaf_density | leaf_fragmentation
---------+------------+------------+---------------+----------------+------------+-------------+---------------+------------------+--------------------
2 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | NaN | NaN
2 | 0 | 1 | 0 | 0 | 0 | 0 | 0 | NaN | NaN
(1 row)
select * from pgstatindex('test_pkey'::text);
select version, tree_level,
index_size / current_setting('block_size')::int as index_size,
root_block_no, internal_pages, leaf_pages, empty_pages, deleted_pages,
avg_leaf_density, leaf_fragmentation
from pgstatindex('test_pkey'::text);
version | tree_level | index_size | root_block_no | internal_pages | leaf_pages | empty_pages | deleted_pages | avg_leaf_density | leaf_fragmentation
---------+------------+------------+---------------+----------------+------------+-------------+---------------+------------------+--------------------
2 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | NaN | NaN
2 | 0 | 1 | 0 | 0 | 0 | 0 | 0 | NaN | NaN
(1 row)
select * from pgstatindex('test_pkey'::name);
select version, tree_level,
index_size / current_setting('block_size')::int as index_size,
root_block_no, internal_pages, leaf_pages, empty_pages, deleted_pages,
avg_leaf_density, leaf_fragmentation
from pgstatindex('test_pkey'::name);
version | tree_level | index_size | root_block_no | internal_pages | leaf_pages | empty_pages | deleted_pages | avg_leaf_density | leaf_fragmentation
---------+------------+------------+---------------+----------------+------------+-------------+---------------+------------------+--------------------
2 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | NaN | NaN
2 | 0 | 1 | 0 | 0 | 0 | 0 | 0 | NaN | NaN
(1 row)
select * from pgstatindex('test_pkey'::regclass);
select version, tree_level,
index_size / current_setting('block_size')::int as index_size,
root_block_no, internal_pages, leaf_pages, empty_pages, deleted_pages,
avg_leaf_density, leaf_fragmentation
from pgstatindex('test_pkey'::regclass);
version | tree_level | index_size | root_block_no | internal_pages | leaf_pages | empty_pages | deleted_pages | avg_leaf_density | leaf_fragmentation
---------+------------+------------+---------------+----------------+------------+-------------+---------------+------------------+--------------------
2 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | NaN | NaN
(1 row)
select pgstatindex(oid) from pg_class where relname = 'test_pkey';
pgstatindex
---------------------------
(2,0,0,0,0,0,0,0,NaN,NaN)
(1 row)
select pgstatindex(relname) from pg_class where relname = 'test_pkey';
pgstatindex
---------------------------
(2,0,0,0,0,0,0,0,NaN,NaN)
2 | 0 | 1 | 0 | 0 | 0 | 0 | 0 | NaN | NaN
(1 row)
select pg_relpages('test');

View File

@ -79,7 +79,6 @@ typedef struct BTIndexStat
uint32 level;
BlockNumber root_blkno;
uint64 root_pages;
uint64 internal_pages;
uint64 leaf_pages;
uint64 empty_pages;
@ -185,7 +184,6 @@ pgstatindex_impl(Relation rel, FunctionCallInfo fcinfo)
}
/* -- init counters -- */
indexStat.root_pages = 0;
indexStat.internal_pages = 0;
indexStat.leaf_pages = 0;
indexStat.empty_pages = 0;
@ -218,7 +216,11 @@ pgstatindex_impl(Relation rel, FunctionCallInfo fcinfo)
/* Determine page type, and update totals */
if (P_ISLEAF(opaque))
if (P_ISDELETED(opaque))
indexStat.deleted_pages++;
else if (P_IGNORE(opaque))
indexStat.empty_pages++; /* this is the "half dead" state */
else if (P_ISLEAF(opaque))
{
int max_avail;
@ -235,12 +237,6 @@ pgstatindex_impl(Relation rel, FunctionCallInfo fcinfo)
if (opaque->btpo_next != P_NONE && opaque->btpo_next < blkno)
indexStat.fragments++;
}
else if (P_ISDELETED(opaque))
indexStat.deleted_pages++;
else if (P_IGNORE(opaque))
indexStat.empty_pages++;
else if (P_ISROOT(opaque))
indexStat.root_pages++;
else
indexStat.internal_pages++;
@ -269,7 +265,7 @@ pgstatindex_impl(Relation rel, FunctionCallInfo fcinfo)
values[j++] = psprintf("%d", indexStat.version);
values[j++] = psprintf("%d", indexStat.level);
values[j++] = psprintf(INT64_FORMAT,
(indexStat.root_pages +
(1 + /* include the metapage in index_size */
indexStat.leaf_pages +
indexStat.internal_pages +
indexStat.deleted_pages +

View File

@ -15,12 +15,26 @@ select * from pgstattuple('test'::regclass);
select pgstattuple(oid) from pg_class where relname = 'test';
select pgstattuple(relname) from pg_class where relname = 'test';
select * from pgstatindex('test_pkey');
select * from pgstatindex('test_pkey'::text);
select * from pgstatindex('test_pkey'::name);
select * from pgstatindex('test_pkey'::regclass);
select pgstatindex(oid) from pg_class where relname = 'test_pkey';
select pgstatindex(relname) from pg_class where relname = 'test_pkey';
select version, tree_level,
index_size / current_setting('block_size')::int as index_size,
root_block_no, internal_pages, leaf_pages, empty_pages, deleted_pages,
avg_leaf_density, leaf_fragmentation
from pgstatindex('test_pkey');
select version, tree_level,
index_size / current_setting('block_size')::int as index_size,
root_block_no, internal_pages, leaf_pages, empty_pages, deleted_pages,
avg_leaf_density, leaf_fragmentation
from pgstatindex('test_pkey'::text);
select version, tree_level,
index_size / current_setting('block_size')::int as index_size,
root_block_no, internal_pages, leaf_pages, empty_pages, deleted_pages,
avg_leaf_density, leaf_fragmentation
from pgstatindex('test_pkey'::name);
select version, tree_level,
index_size / current_setting('block_size')::int as index_size,
root_block_no, internal_pages, leaf_pages, empty_pages, deleted_pages,
avg_leaf_density, leaf_fragmentation
from pgstatindex('test_pkey'::regclass);
select pg_relpages('test');
select pg_relpages('test_pkey');

View File

@ -130,9 +130,9 @@ free_percent | 1.95
<listitem>
<para>
This is the same as <function>pgstattuple(regclass)</function>, except
that the target relation is specified by TEXT. This function is kept
that the target relation is specified as TEXT. This function is kept
because of backward-compatibility so far, and will be deprecated in
the future release.
some future release.
</para>
</listitem>
</varlistentry>
@ -154,13 +154,13 @@ test=&gt; SELECT * FROM pgstatindex('pg_cast_oid_index');
-[ RECORD 1 ]------+------
version | 2
tree_level | 0
index_size | 8192
index_size | 16384
root_block_no | 1
internal_pages | 0
leaf_pages | 1
empty_pages | 0
deleted_pages | 0
avg_leaf_density | 50.27
avg_leaf_density | 54.27
leaf_fragmentation | 0
</programlisting>
</para>
@ -194,13 +194,13 @@ leaf_fragmentation | 0
<row>
<entry><structfield>index_size</structfield></entry>
<entry><type>bigint</type></entry>
<entry>Total number of pages in index</entry>
<entry>Total index size in bytes</entry>
</row>
<row>
<entry><structfield>root_block_no</structfield></entry>
<entry><type>bigint</type></entry>
<entry>Location of root block</entry>
<entry>Location of root page (zero if none)</entry>
</row>
<row>
@ -244,6 +244,13 @@ leaf_fragmentation | 0
</informaltable>
</para>
<para>
The reported <literal>index_size</> will normally correspond to one more
page than is accounted for by <literal>internal_pages + leaf_pages +
empty_pages + deleted_pages</literal>, because it also includes the
index's metapage.
</para>
<para>
As with <function>pgstattuple</>, the results are accumulated
page-by-page, and should not be expected to represent an
@ -260,9 +267,9 @@ leaf_fragmentation | 0
<listitem>
<para>
This is the same as <function>pgstatindex(regclass)</function>, except
that the target index is specified by TEXT. This function is kept
that the target index is specified as TEXT. This function is kept
because of backward-compatibility so far, and will be deprecated in
the future release.
some future release.
</para>
</listitem>
</varlistentry>
@ -351,9 +358,9 @@ pending_tuples | 0
<listitem>
<para>
This is the same as <function>pg_relpages(regclass)</function>, except
that the target relation is specified by TEXT. This function is kept
that the target relation is specified as TEXT. This function is kept
because of backward-compatibility so far, and will be deprecated in
the future release.
some future release.
</para>
</listitem>
</varlistentry>
@ -370,7 +377,7 @@ pending_tuples | 0
<para>
<function>pgstattuple_approx</function> is a faster alternative to
<function>pgstattuple</function> that returns approximate results.
The argument is the target relation's OID.
The argument is the target relation's name or OID.
For example:
<programlisting>
test=&gt; SELECT * FROM pgstattuple_approx('pg_catalog.pg_proc'::regclass);