database/postgres
1
0
Fork 0
mirror of https://github.com/postgres/postgres.git synced 2025-10-22 14:32:25 +03:00
Code Activity

Optimize nbtree backward scan boundary cases.

Browse Source 
Teach _bt_binsrch (and related helper routines like _bt_search and
_bt_compare) about the initial positioning requirements of backward
scans.  Routines like _bt_binsrch already know all about "nextkey"
searches, so it seems natural to teach them about "goback"/backward
searches, too.  These concepts are closely related, and are much easier
to understand when discussed together.

Now that certain implementation details are hidden from _bt_first, it's
straightforward to add a new optimization: backward scans using the <
strategy now avoid extra leaf page accesses in certain "boundary cases".
Consider the following example, which uses the tenk1 table (and its
tenk1_hundred index) from the standard regression tests:

SELECT * FROM tenk1 WHERE hundred < 12 ORDER BY hundred DESC LIMIT 1;

Before this commit, nbtree would scan two leaf pages, even though it was
only really necessary to scan one leaf page.  We'll now descend straight
to the leaf page containing a (12, -inf) high key instead.  The scan
will locate matching non-pivot tuples with "hundred" values starting
from the value 11.  The scan won't waste a page access on the right
sibling leaf page, which cannot possibly contain any matching tuples.

You can think of the optimization added by this commit as disabling an
optimization (the _bt_compare "!pivotsearch" behavior that was added to
Postgres 12 in commit dd299df8) for a small subset of cases where it was
always counterproductive.

Equivalently, you can think of the new optimization as extending the
"pivotsearch" behavior that page deletion by VACUUM has long required
(since the aforementioned Postgres 12 commit went in) to other, similar
cases.  Obviously, this isn't strictly necessary for these new cases
(unlike VACUUM, _bt_first is prepared to move the scan to the left once
on the leaf level), but the underlying principle is the same.

Author: Peter Geoghegan <pg@bowt.ie>
Reviewed-By: Matthias van de Meent <boekewurm+postgres@gmail.com>
Discussion: https://postgr.es/m/CAH2-Wz=XPzM8HzaLPq278Vms420mVSHfgs9wi5tjFKHcapZCEw@mail.gmail.com
This commit is contained in:
Peter Geoghegan
2023-12-08 11:05:17 -08:00
parent b437571714
commit c9c0589fda
7 changed files with 210 additions and 149 deletions
Download Patch File Download Diff File Expand all files Collapse all files

11
src/include/access/nbtree.h
View File

@@ -763,13 +763,8 @@ typedef BTStackData *BTStack;
* bit, but may not when inserting into an INCLUDE index (tuple header value
* is affected by the NULL-ness of both key and non-key attributes).
*
* When nextkey is false (the usual case), _bt_search and _bt_binsrch will
* locate the first item >= scankey. When nextkey is true, they will locate
* the first item > scan key.
*
* pivotsearch is set to true by callers that want to re-find a leaf page
* using a scankey built from a leaf page's high key. Most callers set this
* to false.
* See comments in _bt_first for an explanation of the nextkey and backward
* fields.
*
* scantid is the heap TID that is used as a final tiebreaker attribute. It
* is set to NULL when index scan doesn't need to find a position for a
@@ -792,7 +787,7 @@ typedef struct BTScanInsertData
bool allequalimage;
bool anynullkeys;
bool nextkey;
bool pivotsearch;
bool backward; /* backward index scan? */
ItemPointer scantid; /* tiebreaker for scankeys */
int keysz; /* Size of scankeys array */
ScanKeyData scankeys[INDEX_MAX_KEYS]; /* Must appear last */