analyze.c

来自「PostgreSQL 8.1.4的源码 适用于Linux下的开源数据库系统」· C语言 代码 · 共 2,242 行 · 第 1/5 页

/*-------------------------------------------------------------------------
 *
 * analyze.c
 *	  the Postgres statistics generator
 *
 * Portions Copyright (c) 1996-2005, PostgreSQL Global Development Group
 * Portions Copyright (c) 1994, Regents of the University of California
 *
 *
 * IDENTIFICATION
 *	  $PostgreSQL: pgsql/src/backend/commands/analyze.c,v 1.89.2.1 2005/11/22 18:23:06 momjian Exp $
 *
 *-------------------------------------------------------------------------
 */
#include "postgres.h"

#include <math.h>

#include "access/heapam.h"
#include "access/tuptoaster.h"
#include "catalog/catalog.h"
#include "catalog/index.h"
#include "catalog/indexing.h"
#include "catalog/namespace.h"
#include "catalog/pg_operator.h"
#include "commands/vacuum.h"
#include "executor/executor.h"
#include "miscadmin.h"
#include "parser/parse_expr.h"
#include "parser/parse_oper.h"
#include "parser/parse_relation.h"
#include "pgstat.h"
#include "utils/acl.h"
#include "utils/builtins.h"
#include "utils/datum.h"
#include "utils/fmgroids.h"
#include "utils/lsyscache.h"
#include "utils/memutils.h"
#include "utils/syscache.h"
#include "utils/tuplesort.h"


/* Data structure for Algorithm S from Knuth 3.4.2 */
typedef struct
{
	BlockNumber N;				/* number of blocks, known in advance */
	int			n;				/* desired sample size */
	BlockNumber t;				/* current block number */
	int			m;				/* blocks selected so far */
} BlockSamplerData;
typedef BlockSamplerData *BlockSampler;

/* Per-index data for ANALYZE */
typedef struct AnlIndexData
{
	IndexInfo  *indexInfo;		/* BuildIndexInfo result */
	double		tupleFract;		/* fraction of rows for partial index */
	VacAttrStats **vacattrstats;	/* index attrs to analyze */
	int			attr_cnt;
} AnlIndexData;


/* Default statistics target (GUC parameter) */
int			default_statistics_target = 10;

static int	elevel = -1;

static MemoryContext anl_context = NULL;


static void BlockSampler_Init(BlockSampler bs, BlockNumber nblocks,
				  int samplesize);
static bool BlockSampler_HasMore(BlockSampler bs);
static BlockNumber BlockSampler_Next(BlockSampler bs);
static void compute_index_stats(Relation onerel, double totalrows,
					AnlIndexData *indexdata, int nindexes,
					HeapTuple *rows, int numrows,
					MemoryContext col_context);
static VacAttrStats *examine_attribute(Relation onerel, int attnum);
static int acquire_sample_rows(Relation onerel, HeapTuple *rows,
					int targrows, double *totalrows, double *totaldeadrows);
static double random_fract(void);
static double init_selection_state(int n);
static double get_next_S(double t, int n, double *stateptr);
static int	compare_rows(const void *a, const void *b);
static void update_attstats(Oid relid, int natts, VacAttrStats **vacattrstats);
static Datum std_fetch_func(VacAttrStatsP stats, int rownum, bool *isNull);
static Datum ind_fetch_func(VacAttrStatsP stats, int rownum, bool *isNull);

static bool std_typanalyze(VacAttrStats *stats);


/*
 *	analyze_rel() -- analyze one relation
 */
void
analyze_rel(Oid relid, VacuumStmt *vacstmt)
{
	Relation	onerel;
	int			attr_cnt,
				tcnt,
				i,
				ind;
	Relation   *Irel;
	int			nindexes;
	bool		hasindex;
	bool		analyzableindex;
	VacAttrStats **vacattrstats;
	AnlIndexData *indexdata;
	int			targrows,
				numrows;
	double		totalrows,
				totaldeadrows;
	HeapTuple  *rows;

	if (vacstmt->verbose)
		elevel = INFO;
	else
		elevel = DEBUG2;

	/*
	 * Use the current context for storing analysis info.  vacuum.c ensures
	 * that this context will be cleared when I return, thus releasing the
	 * memory allocated here.
	 */
	anl_context = CurrentMemoryContext;

	/*
	 * Check for user-requested abort.	Note we want this to be inside a
	 * transaction, so xact.c doesn't issue useless WARNING.
	 */
	CHECK_FOR_INTERRUPTS();

	/*
	 * Race condition -- if the pg_class tuple has gone away since the last
	 * time we saw it, we don't need to process it.
	 */
	if (!SearchSysCacheExists(RELOID,
							  ObjectIdGetDatum(relid),
							  0, 0, 0))
		return;

	/*
	 * Open the class, getting only a read lock on it, and check permissions.
	 * Permissions check should match vacuum's check!
	 */
	onerel = relation_open(relid, AccessShareLock);

	if (!(pg_class_ownercheck(RelationGetRelid(onerel), GetUserId()) ||
		  (pg_database_ownercheck(MyDatabaseId, GetUserId()) && !onerel->rd_rel->relisshared)))
	{
		/* No need for a WARNING if we already complained during VACUUM */
		if (!vacstmt->vacuum)
			ereport(WARNING,
					(errmsg("skipping \"%s\" --- only table or database owner can analyze it",
							RelationGetRelationName(onerel))));
		relation_close(onerel, AccessShareLock);
		return;
	}

	/*
	 * Check that it's a plain table; we used to do this in get_rel_oids() but
	 * seems safer to check after we've locked the relation.
	 */
	if (onerel->rd_rel->relkind != RELKIND_RELATION)
	{
		/* No need for a WARNING if we already complained during VACUUM */
		if (!vacstmt->vacuum)
			ereport(WARNING,
					(errmsg("skipping \"%s\" --- cannot analyze indexes, views, or special system tables",
							RelationGetRelationName(onerel))));
		relation_close(onerel, AccessShareLock);
		return;
	}

	/*
	 * Silently ignore tables that are temp tables of other backends ---
	 * trying to analyze these is rather pointless, since their contents are
	 * probably not up-to-date on disk.  (We don't throw a warning here; it
	 * would just lead to chatter during a database-wide ANALYZE.)
	 */
	if (isOtherTempNamespace(RelationGetNamespace(onerel)))
	{
		relation_close(onerel, AccessShareLock);
		return;
	}

	/*
	 * We can ANALYZE any table except pg_statistic. See update_attstats
	 */
	if (RelationGetRelid(onerel) == StatisticRelationId)
	{
		relation_close(onerel, AccessShareLock);
		return;
	}

	ereport(elevel,
			(errmsg("analyzing \"%s.%s\"",
					get_namespace_name(RelationGetNamespace(onerel)),
					RelationGetRelationName(onerel))));

	/*
	 * Determine which columns to analyze
	 *
	 * Note that system attributes are never analyzed.
	 */
	if (vacstmt->va_cols != NIL)
	{
		ListCell   *le;

		vacattrstats = (VacAttrStats **) palloc(list_length(vacstmt->va_cols) *
												sizeof(VacAttrStats *));
		tcnt = 0;
		foreach(le, vacstmt->va_cols)
		{
			char	   *col = strVal(lfirst(le));

			i = attnameAttNum(onerel, col, false);
			vacattrstats[tcnt] = examine_attribute(onerel, i);
			if (vacattrstats[tcnt] != NULL)
				tcnt++;
		}
		attr_cnt = tcnt;
	}
	else
	{
		attr_cnt = onerel->rd_att->natts;
		vacattrstats = (VacAttrStats **)
			palloc(attr_cnt * sizeof(VacAttrStats *));
		tcnt = 0;
		for (i = 1; i <= attr_cnt; i++)
		{
			vacattrstats[tcnt] = examine_attribute(onerel, i);
			if (vacattrstats[tcnt] != NULL)
				tcnt++;
		}
		attr_cnt = tcnt;
	}

	/*
	 * Open all indexes of the relation, and see if there are any analyzable
	 * columns in the indexes.	We do not analyze index columns if there was
	 * an explicit column list in the ANALYZE command, however.
	 */
	vac_open_indexes(onerel, AccessShareLock, &nindexes, &Irel);
	hasindex = (nindexes > 0);
	indexdata = NULL;
	analyzableindex = false;
	if (hasindex)
	{
		indexdata = (AnlIndexData *) palloc0(nindexes * sizeof(AnlIndexData));
		for (ind = 0; ind < nindexes; ind++)
		{
			AnlIndexData *thisdata = &indexdata[ind];
			IndexInfo  *indexInfo;

			thisdata->indexInfo = indexInfo = BuildIndexInfo(Irel[ind]);
			thisdata->tupleFract = 1.0; /* fix later if partial */
			if (indexInfo->ii_Expressions != NIL && vacstmt->va_cols == NIL)
			{
				ListCell   *indexpr_item = list_head(indexInfo->ii_Expressions);

				thisdata->vacattrstats = (VacAttrStats **)
					palloc(indexInfo->ii_NumIndexAttrs * sizeof(VacAttrStats *));
				tcnt = 0;
				for (i = 0; i < indexInfo->ii_NumIndexAttrs; i++)
				{
					int			keycol = indexInfo->ii_KeyAttrNumbers[i];

					if (keycol == 0)
					{
						/* Found an index expression */
						Node	   *indexkey;

						if (indexpr_item == NULL)		/* shouldn't happen */
							elog(ERROR, "too few entries in indexprs list");
						indexkey = (Node *) lfirst(indexpr_item);
						indexpr_item = lnext(indexpr_item);

						/*
						 * Can't analyze if the opclass uses a storage type
						 * different from the expression result type. We'd get
						 * confused because the type shown in pg_attribute for
						 * the index column doesn't match what we are getting
						 * from the expression. Perhaps this can be fixed
						 * someday, but for now, punt.
						 */
						if (exprType(indexkey) !=
							Irel[ind]->rd_att->attrs[i]->atttypid)
							continue;

						thisdata->vacattrstats[tcnt] =
							examine_attribute(Irel[ind], i + 1);
						if (thisdata->vacattrstats[tcnt] != NULL)
						{
							tcnt++;
							analyzableindex = true;
						}
					}
				}
				thisdata->attr_cnt = tcnt;
			}
		}
	}

	/*
	 * Quit if no analyzable columns
	 */
	if (attr_cnt <= 0 && !analyzableindex)
	{
		/*
		 * We report that the table is empty; this is just so that the
		 * autovacuum code doesn't go nuts trying to get stats about a
		 * zero-column table.
		 */
		if (!vacstmt->vacuum)
			pgstat_report_analyze(RelationGetRelid(onerel),
								  onerel->rd_rel->relisshared,
								  0, 0);

		vac_close_indexes(nindexes, Irel, AccessShareLock);
		relation_close(onerel, AccessShareLock);
		return;
	}

	/*
	 * Determine how many rows we need to sample, using the worst case from
	 * all analyzable columns.	We use a lower bound of 100 rows to avoid
	 * possible overflow in Vitter's algorithm.
	 */
	targrows = 100;
	for (i = 0; i < attr_cnt; i++)
	{
		if (targrows < vacattrstats[i]->minrows)
			targrows = vacattrstats[i]->minrows;
	}
	for (ind = 0; ind < nindexes; ind++)
	{
		AnlIndexData *thisdata = &indexdata[ind];

		for (i = 0; i < thisdata->attr_cnt; i++)
		{
			if (targrows < thisdata->vacattrstats[i]->minrows)
				targrows = thisdata->vacattrstats[i]->minrows;
		}
	}

	/*
	 * Acquire the sample rows
	 */
	rows = (HeapTuple *) palloc(targrows * sizeof(HeapTuple));
	numrows = acquire_sample_rows(onerel, rows, targrows,
								  &totalrows, &totaldeadrows);

	/*
	 * Compute the statistics.	Temporary results during the calculations for
	 * each column are stored in a child context.  The calc routines are
	 * responsible to make sure that whatever they store into the VacAttrStats
	 * structure is allocated in anl_context.
	 */
	if (numrows > 0)
	{
		MemoryContext col_context,
					old_context;

		col_context = AllocSetContextCreate(anl_context,
											"Analyze Column",
											ALLOCSET_DEFAULT_MINSIZE,
											ALLOCSET_DEFAULT_INITSIZE,
											ALLOCSET_DEFAULT_MAXSIZE);
		old_context = MemoryContextSwitchTo(col_context);

		for (i = 0; i < attr_cnt; i++)
		{
			VacAttrStats *stats = vacattrstats[i];

			stats->rows = rows;
			stats->tupDesc = onerel->rd_att;
			(*stats->compute_stats) (stats,
									 std_fetch_func,
									 numrows,
									 totalrows);
			MemoryContextResetAndDeleteChildren(col_context);
		}

		if (hasindex)
			compute_index_stats(onerel, totalrows,
								indexdata, nindexes,
								rows, numrows,
								col_context);

		MemoryContextSwitchTo(old_context);
		MemoryContextDelete(col_context);

		/*
		 * Emit the completed stats rows into pg_statistic, replacing any
		 * previous statistics for the target columns.	(If there are stats in
		 * pg_statistic for columns we didn't process, we leave them alone.)
		 */
		update_attstats(relid, attr_cnt, vacattrstats);

		for (ind = 0; ind < nindexes; ind++)
		{
			AnlIndexData *thisdata = &indexdata[ind];

			update_attstats(RelationGetRelid(Irel[ind]),
							thisdata->attr_cnt, thisdata->vacattrstats);
		}
	}

	/*
	 * If we are running a standalone ANALYZE, update pages/tuples stats in
	 * pg_class.  We know the accurate page count from the smgr, but only an
	 * approximate number of tuples; therefore, if we are part of VACUUM
	 * ANALYZE do *not* overwrite the accurate count already inserted by
	 * VACUUM.	The same consideration applies to indexes.
	 */
	if (!vacstmt->vacuum)
	{
		vac_update_relstats(RelationGetRelid(onerel),
							RelationGetNumberOfBlocks(onerel),
							totalrows,
							hasindex);
		for (ind = 0; ind < nindexes; ind++)
		{
			AnlIndexData *thisdata = &indexdata[ind];
			double		totalindexrows;

			totalindexrows = ceil(thisdata->tupleFract * totalrows);
			vac_update_relstats(RelationGetRelid(Irel[ind]),
								RelationGetNumberOfBlocks(Irel[ind]),
								totalindexrows,
								false);
		}

		/* report results to the stats collector, too */
		pgstat_report_analyze(RelationGetRelid(onerel),
							  onerel->rd_rel->relisshared,
							  totalrows, totaldeadrows);
	}

	/* Done with indexes */
	vac_close_indexes(nindexes, Irel, NoLock);

	/*
	 * Close source relation now, but keep lock so that no one deletes it
	 * before we commit.  (If someone did, they'd fail to clean up the entries
	 * we made in pg_statistic.)
	 */