pgstatapprox.c File Reference

#include "postgres.h"
#include "access/visibilitymap.h"
#include "access/transam.h"
#include "access/xact.h"
#include "access/multixact.h"
#include "access/htup_details.h"
#include "catalog/namespace.h"
#include "funcapi.h"
#include "miscadmin.h"
#include "storage/bufmgr.h"
#include "storage/freespace.h"
#include "storage/procarray.h"
#include "storage/lmgr.h"
#include "utils/builtins.h"
#include "utils/tqual.h"
#include "commands/vacuum.h"

Include dependency graph for pgstatapprox.c:

Go to the source code of this file.

Data Structures
struct	output_type

Macros
#define	NUM_OUTPUT_COLUMNS   10

Typedefs
typedef struct output_type	output_type

Functions
	PG_FUNCTION_INFO_V1 (pgstattuple_approx)
	PG_FUNCTION_INFO_V1 (pgstattuple_approx_v1_5)
Datum	pgstattuple_approx_internal (Oid relid, FunctionCallInfo fcinfo)
static void	statapprox_heap (Relation rel, output_type *stat)
Datum	pgstattuple_approx (PG_FUNCTION_ARGS)
Datum	pgstattuple_approx_v1_5 (PG_FUNCTION_ARGS)

Macro Definition Documentation

NUM_OUTPUT_COLUMNS

#define NUM_OUTPUT_COLUMNS   10

Definition at line 50 of file pgstatapprox.c.

Referenced by pgstattuple_approx_internal().

Typedef Documentation

output_type

typedef struct output_type output_type

Function Documentation

PG_FUNCTION_INFO_V1() [1/2]

PG_FUNCTION_INFO_V1

(

pgstattuple_approx

)

PG_FUNCTION_INFO_V1() [2/2]

PG_FUNCTION_INFO_V1

(

pgstattuple_approx_v1_5

)

pgstattuple_approx()

Datum pgstattuple_approx

(

PG_FUNCTION_ARGS

)

Definition at line 222 of file pgstatapprox.c.

References ereport, errcode(), errmsg(), ERROR, PG_GETARG_OID, PG_RETURN_DATUM, pgstattuple_approx_internal(), and superuser().

{
    Oid         relid = PG_GETARG_OID(0);

    if (!superuser())
        ereport(ERROR,
                (errcode(ERRCODE_INSUFFICIENT_PRIVILEGE),
                 (errmsg("must be superuser to use pgstattuple functions"))));

    PG_RETURN_DATUM(pgstattuple_approx_internal(relid, fcinfo));
}

pgstattuple_approx_internal()

Datum pgstattuple_approx_internal	(	Oid	relid,
		FunctionCallInfo	fcinfo
	)

Definition at line 250 of file pgstatapprox.c.

References AccessShareLock, output_type::dead_tuple_count, output_type::dead_tuple_len, output_type::dead_tuple_percent, elog, ereport, errcode(), errmsg(), ERROR, Float8GetDatum(), output_type::free_percent, output_type::free_space, get_call_result_type(), heap_form_tuple(), HeapTupleGetDatum, i, Int64GetDatum(), tupleDesc::natts, NUM_OUTPUT_COLUMNS, RelationData::rd_rel, relation_close(), RELATION_IS_OTHER_TEMP, relation_open(), RelationGetRelationName, output_type::scanned_percent, stat, statapprox_heap(), output_type::table_len, output_type::tuple_count, output_type::tuple_len, output_type::tuple_percent, TYPEFUNC_COMPOSITE, and values.

Referenced by pgstattuple_approx(), and pgstattuple_approx_v1_5().

{
    Relation    rel;
    output_type stat = {0};
    TupleDesc   tupdesc;
    bool        nulls[NUM_OUTPUT_COLUMNS];
    Datum       values[NUM_OUTPUT_COLUMNS];
    HeapTuple   ret;
    int         i = 0;

    if (get_call_result_type(fcinfo, NULL, &tupdesc) != TYPEFUNC_COMPOSITE)
        elog(ERROR, "return type must be a row type");

    if (tupdesc->natts != NUM_OUTPUT_COLUMNS)
        elog(ERROR, "incorrect number of output arguments");

    rel = relation_open(relid, AccessShareLock);

    /*
     * Reject attempts to read non-local temporary relations; we would be
     * likely to get wrong data since we have no visibility into the owning
     * session's local buffers.
     */
    if (RELATION_IS_OTHER_TEMP(rel))
        ereport(ERROR,
                (errcode(ERRCODE_FEATURE_NOT_SUPPORTED),
                 errmsg("cannot access temporary tables of other sessions")));

    /*
     * We support only ordinary relations and materialised views, because we
     * depend on the visibility map and free space map for our estimates about
     * unscanned pages.
     */
    if (!(rel->rd_rel->relkind == RELKIND_RELATION ||
          rel->rd_rel->relkind == RELKIND_MATVIEW))
        ereport(ERROR,
                (errcode(ERRCODE_FEATURE_NOT_SUPPORTED),
                 errmsg("\"%s\" is not a table or materialized view",
                        RelationGetRelationName(rel))));

    statapprox_heap(rel, &stat);

    relation_close(rel, AccessShareLock);

    memset(nulls, 0, sizeof(nulls));

    values[i++] = Int64GetDatum(stat.table_len);
    values[i++] = Float8GetDatum(stat.scanned_percent);
    values[i++] = Int64GetDatum(stat.tuple_count);
    values[i++] = Int64GetDatum(stat.tuple_len);
    values[i++] = Float8GetDatum(stat.tuple_percent);
    values[i++] = Int64GetDatum(stat.dead_tuple_count);
    values[i++] = Int64GetDatum(stat.dead_tuple_len);
    values[i++] = Float8GetDatum(stat.dead_tuple_percent);
    values[i++] = Int64GetDatum(stat.free_space);
    values[i++] = Float8GetDatum(stat.free_percent);

    ret = heap_form_tuple(tupdesc, values, nulls);
    return HeapTupleGetDatum(ret);
}

pgstattuple_approx_v1_5()

Datum pgstattuple_approx_v1_5

(

PG_FUNCTION_ARGS

)

Definition at line 242 of file pgstatapprox.c.

References PG_GETARG_OID, PG_RETURN_DATUM, and pgstattuple_approx_internal().

{
    Oid         relid = PG_GETARG_OID(0);

    PG_RETURN_DATUM(pgstattuple_approx_internal(relid, fcinfo));
}

statapprox_heap()

static void statapprox_heap ( Relation  rel, output_type *  stat  )

static

Definition at line 63 of file pgstatapprox.c.

References Assert, BAS_BULKREAD, buf, BUFFER_LOCK_SHARE, BufferGetPage, BufferIsValid, CHECK_FOR_INTERRUPTS, output_type::dead_tuple_count, output_type::dead_tuple_len, output_type::dead_tuple_percent, elog, ERROR, FirstOffsetNumber, output_type::free_percent, output_type::free_space, GetAccessStrategy(), GetOldestXmin(), GetRecordedFreeSpace(), HEAPTUPLE_DEAD, HEAPTUPLE_DELETE_IN_PROGRESS, HEAPTUPLE_INSERT_IN_PROGRESS, HEAPTUPLE_LIVE, HEAPTUPLE_RECENTLY_DEAD, HeapTupleSatisfiesVacuum(), InvalidBuffer, ItemIdGetLength, ItemIdIsDead, ItemIdIsNormal, ItemIdIsRedirected, ItemIdIsUsed, ItemPointerSet, LockBuffer(), MAIN_FORKNUM, OffsetNumberNext, OldestXmin, PageGetHeapFreeSpace(), PageGetItem, PageGetItemId, PageGetMaxOffsetNumber, PageIsEmpty, PageIsNew, PROCARRAY_FLAGS_VACUUM, RBM_NORMAL, ReadBufferExtended(), RelationGetNumberOfBlocks, RelationGetRelid, ReleaseBuffer(), output_type::scanned_percent, SizeOfPageHeaderData, HeapTupleData::t_data, HeapTupleData::t_len, HeapTupleData::t_self, HeapTupleData::t_tableOid, output_type::table_len, output_type::tuple_count, output_type::tuple_len, output_type::tuple_percent, UnlockReleaseBuffer(), vac_estimate_reltuples(), and VM_ALL_VISIBLE.

Referenced by pgstattuple_approx_internal().

{
    BlockNumber scanned,
                nblocks,
                blkno;
    Buffer      vmbuffer = InvalidBuffer;
    BufferAccessStrategy bstrategy;
    TransactionId OldestXmin;

    OldestXmin = GetOldestXmin(rel, PROCARRAY_FLAGS_VACUUM);
    bstrategy = GetAccessStrategy(BAS_BULKREAD);

    nblocks = RelationGetNumberOfBlocks(rel);
    scanned = 0;

    for (blkno = 0; blkno < nblocks; blkno++)
    {
        Buffer      buf;
        Page        page;
        OffsetNumber offnum,
                    maxoff;
        Size        freespace;

        CHECK_FOR_INTERRUPTS();

        /*
         * If the page has only visible tuples, then we can find out the free
         * space from the FSM and move on.
         */
        if (VM_ALL_VISIBLE(rel, blkno, &vmbuffer))
        {
            freespace = GetRecordedFreeSpace(rel, blkno);
            stat->tuple_len += BLCKSZ - freespace;
            stat->free_space += freespace;
            continue;
        }

        buf = ReadBufferExtended(rel, MAIN_FORKNUM, blkno,
                                 RBM_NORMAL, bstrategy);

        LockBuffer(buf, BUFFER_LOCK_SHARE);

        page = BufferGetPage(buf);

        /*
         * It's not safe to call PageGetHeapFreeSpace() on new pages, so we
         * treat them as being free space for our purposes.
         */
        if (!PageIsNew(page))
            stat->free_space += PageGetHeapFreeSpace(page);
        else
            stat->free_space += BLCKSZ - SizeOfPageHeaderData;

        /* We may count the page as scanned even if it's new/empty */
        scanned++;

        if (PageIsNew(page) || PageIsEmpty(page))
        {
            UnlockReleaseBuffer(buf);
            continue;
        }

        /*
         * Look at each tuple on the page and decide whether it's live or
         * dead, then count it and its size. Unlike lazy_scan_heap, we can
         * afford to ignore problems and special cases.
         */
        maxoff = PageGetMaxOffsetNumber(page);

        for (offnum = FirstOffsetNumber;
             offnum <= maxoff;
             offnum = OffsetNumberNext(offnum))
        {
            ItemId      itemid;
            HeapTupleData tuple;

            itemid = PageGetItemId(page, offnum);

            if (!ItemIdIsUsed(itemid) || ItemIdIsRedirected(itemid) ||
                ItemIdIsDead(itemid))
            {
                continue;
            }

            Assert(ItemIdIsNormal(itemid));

            ItemPointerSet(&(tuple.t_self), blkno, offnum);

            tuple.t_data = (HeapTupleHeader) PageGetItem(page, itemid);
            tuple.t_len = ItemIdGetLength(itemid);
            tuple.t_tableOid = RelationGetRelid(rel);

            /*
             * We follow VACUUM's lead in counting INSERT_IN_PROGRESS tuples
             * as "dead" while DELETE_IN_PROGRESS tuples are "live".  We don't
             * bother distinguishing tuples inserted/deleted by our own
             * transaction.
             */
            switch (HeapTupleSatisfiesVacuum(&tuple, OldestXmin, buf))
            {
                case HEAPTUPLE_LIVE:
                case HEAPTUPLE_DELETE_IN_PROGRESS:
                    stat->tuple_len += tuple.t_len;
                    stat->tuple_count++;
                    break;
                case HEAPTUPLE_DEAD:
                case HEAPTUPLE_RECENTLY_DEAD:
                case HEAPTUPLE_INSERT_IN_PROGRESS:
                    stat->dead_tuple_len += tuple.t_len;
                    stat->dead_tuple_count++;
                    break;
                default:
                    elog(ERROR, "unexpected HeapTupleSatisfiesVacuum result");
                    break;
            }
        }

        UnlockReleaseBuffer(buf);
    }

    stat->table_len = (uint64) nblocks * BLCKSZ;

    /*
     * We don't know how many tuples are in the pages we didn't scan, so
     * extrapolate the live-tuple count to the whole table in the same way
     * that VACUUM does.  (Like VACUUM, we're not taking a random sample, so
     * just extrapolating linearly seems unsafe.)  There should be no dead
     * tuples in all-visible pages, so no correction is needed for that, and
     * we already accounted for the space in those pages, too.
     */
    stat->tuple_count = vac_estimate_reltuples(rel, nblocks, scanned,
                                               stat->tuple_count);

    /*
     * Calculate percentages if the relation has one or more pages.
     */
    if (nblocks != 0)
    {
        stat->scanned_percent = 100 * scanned / nblocks;
        stat->tuple_percent = 100.0 * stat->tuple_len / stat->table_len;
        stat->dead_tuple_percent = 100.0 * stat->dead_tuple_len / stat->table_len;
        stat->free_percent = 100.0 * stat->free_space / stat->table_len;
    }

    if (BufferIsValid(vmbuffer))
    {
        ReleaseBuffer(vmbuffer);
        vmbuffer = InvalidBuffer;
    }
}