]> granicus.if.org Git - postgresql/blobdiff - src/include/pgstat.h
Fix up pgstats counting of live and dead tuples to recognize that committed
[postgresql] / src / include / pgstat.h
index 0a943590ec737ef18563a5a5a704195832d6044d..476fd47dc7b0ab48fba50abc3b4ad3e965c522ae 100644 (file)
@@ -3,9 +3,9 @@
  *
  *     Definitions for the PostgreSQL statistics collector daemon.
  *
- *     Copyright (c) 2001-2005, PostgreSQL Global Development Group
+ *     Copyright (c) 2001-2007, PostgreSQL Global Development Group
  *
- *     $PostgreSQL: pgsql/src/include/pgstat.h,v 1.41 2006/01/18 20:35:06 tgl Exp $
+ *     $PostgreSQL: pgsql/src/include/pgstat.h,v 1.59 2007/05/27 03:50:39 tgl Exp $
  * ----------
  */
 #ifndef PGSTAT_H
 #include "utils/rel.h"
 #include "utils/timestamp.h"
 
+
 /* ----------
- * The types of backend/postmaster -> collector messages
+ * The types of backend -> collector messages
  * ----------
  */
 typedef enum StatMsgType
 {
        PGSTAT_MTYPE_DUMMY,
-       PGSTAT_MTYPE_BESTART,
-       PGSTAT_MTYPE_BETERM,
-       PGSTAT_MTYPE_ACTIVITY,
        PGSTAT_MTYPE_TABSTAT,
        PGSTAT_MTYPE_TABPURGE,
        PGSTAT_MTYPE_DROPDB,
        PGSTAT_MTYPE_RESETCOUNTER,
        PGSTAT_MTYPE_AUTOVAC_START,
        PGSTAT_MTYPE_VACUUM,
-       PGSTAT_MTYPE_ANALYZE
+       PGSTAT_MTYPE_ANALYZE,
+       PGSTAT_MTYPE_BGWRITER
 } StatMsgType;
 
 /* ----------
@@ -41,35 +40,13 @@ typedef enum StatMsgType
  */
 typedef int64 PgStat_Counter;
 
-
-/* ------------------------------------------------------------
- * Message formats follow
- * ------------------------------------------------------------
- */
-
-
-/* ----------
- * PgStat_MsgHdr                               The common message header
- * ----------
- */
-typedef struct PgStat_MsgHdr
-{
-       StatMsgType m_type;
-       int                     m_size;
-       int                     m_backendid;
-       int                     m_procpid;
-} PgStat_MsgHdr;
-
-/* ----------
- * Space available in a message.  This will keep the UDP packets below 1K,
- * which should fit unfragmented into the MTU of the lo interface on most
- * platforms. Does anybody care for platforms where it doesn't?
- * ----------
- */
-#define PGSTAT_MSG_PAYLOAD     (1000 - sizeof(PgStat_MsgHdr))
-
 /* ----------
- * PgStat_TableEntry                   Per-table info in a MsgTabstat
+ * PgStat_TableCounts                  The actual per-table counts kept by a backend
+ *
+ * This struct should contain only actual event counters, because we memcmp
+ * it against zeroes to detect whether there are any counts to transmit.
+ * It is a component of PgStat_TableStatus (within-backend state) and
+ * PgStat_TableEntry (the transmitted message format).
  *
  * Note: for a table, tuples_returned is the number of tuples successfully
  * fetched by heap_getnext, while tuples_fetched is the number of tuples
@@ -77,12 +54,14 @@ typedef struct PgStat_MsgHdr
  * For an index, tuples_returned is the number of index entries returned by
  * the index AM, while tuples_fetched is the number of tuples successfully
  * fetched by heap_fetch under the control of simple indexscans for this index.
+ *
+ * tuples_inserted/tuples_updated/tuples_deleted count attempted actions,
+ * regardless of whether the transaction committed.  new_live_tuples and
+ * new_dead_tuples are properly adjusted depending on commit or abort.
  * ----------
  */
-typedef struct PgStat_TableEntry
+typedef struct PgStat_TableCounts
 {
-       Oid                     t_id;
-
        PgStat_Counter t_numscans;
 
        PgStat_Counter t_tuples_returned;
@@ -92,95 +71,104 @@ typedef struct PgStat_TableEntry
        PgStat_Counter t_tuples_updated;
        PgStat_Counter t_tuples_deleted;
 
+       PgStat_Counter t_new_live_tuples;
+       PgStat_Counter t_new_dead_tuples;
+
        PgStat_Counter t_blocks_fetched;
        PgStat_Counter t_blocks_hit;
-} PgStat_TableEntry;
+} PgStat_TableCounts;
 
 
-/* ----------
- * PgStat_MsgDummy                             A dummy message, ignored by the collector
- * ----------
+/* ------------------------------------------------------------
+ * Structures kept in backend local memory while accumulating counts
+ * ------------------------------------------------------------
  */
-typedef struct PgStat_MsgDummy
-{
-       PgStat_MsgHdr m_hdr;
-       char            m_dummy[512];
-} PgStat_MsgDummy;
+
 
 /* ----------
- * PgStat_MsgBestart                   Sent by the backend on startup
+ * PgStat_TableStatus                  Per-table status within a backend
+ *
+ * Most of the event counters are nontransactional, ie, we count events
+ * in committed and aborted transactions alike.  For these, we just count
+ * directly in the PgStat_TableStatus.  However, new_live_tuples and
+ * new_dead_tuples must be derived from tuple insertion and deletion counts
+ * with awareness of whether the transaction or subtransaction committed or
+ * aborted.  Hence, we also keep a stack of per-(sub)transaction status
+ * records for every table modified in the current transaction.  At commit
+ * or abort, we propagate tuples_inserted and tuples_deleted up to the
+ * parent subtransaction level, or out to the parent PgStat_TableStatus,
+ * as appropriate.
  * ----------
  */
-typedef struct PgStat_MsgBestart
+typedef struct PgStat_TableStatus
 {
-       PgStat_MsgHdr m_hdr;
-       Oid                     m_databaseid;
-       Oid                     m_userid;
-       SockAddr        m_clientaddr;
-} PgStat_MsgBestart;
+       Oid                     t_id;                           /* table's OID */
+       bool            t_shared;                       /* is it a shared catalog? */
+       struct PgStat_TableXactStatus *trans;   /* lowest subxact's counts */
+       PgStat_TableCounts t_counts;    /* event counts to be sent */
+} PgStat_TableStatus;
 
 /* ----------
- * PgStat_MsgBeterm                            Sent by the postmaster after backend exit
+ * PgStat_TableXactStatus              Per-table, per-subtransaction status
  * ----------
  */
-typedef struct PgStat_MsgBeterm
+typedef struct PgStat_TableXactStatus
 {
-       PgStat_MsgHdr m_hdr;
-} PgStat_MsgBeterm;
+       PgStat_Counter tuples_inserted; /* tuples inserted in (sub)xact */
+       PgStat_Counter tuples_deleted;  /* tuples deleted in (sub)xact */
+       int                     nest_level;                     /* subtransaction nest level */
+       /* links to other structs for same relation: */
+       struct PgStat_TableXactStatus *upper;   /* next higher subxact if any */
+       PgStat_TableStatus *parent;                             /* per-table status */
+       /* structs of same subxact level are linked here: */
+       struct PgStat_TableXactStatus *next;    /* next of same subxact */
+} PgStat_TableXactStatus;
+
+
+/* ------------------------------------------------------------
+ * Message formats follow
+ * ------------------------------------------------------------
+ */
+
 
 /* ----------
- * PgStat_MsgAutovacStart              Sent by the autovacuum daemon to signal
- *                                                             that a database is going to be processed
+ * PgStat_MsgHdr                               The common message header
  * ----------
  */
-typedef struct PgStat_MsgAutovacStart
+typedef struct PgStat_MsgHdr
 {
-       PgStat_MsgHdr m_hdr;
-       Oid                     m_databaseid;
-       TimestampTz m_start_time;
-} PgStat_MsgAutovacStart;
+       StatMsgType m_type;
+       int                     m_size;
+} PgStat_MsgHdr;
 
 /* ----------
- * PgStat_MsgVacuum                            Sent by the backend or autovacuum daemon
- *                                                             after VACUUM or VACUUM ANALYZE
+ * Space available in a message.  This will keep the UDP packets below 1K,
+ * which should fit unfragmented into the MTU of the lo interface on most
+ * platforms. Does anybody care for platforms where it doesn't?
  * ----------
  */
-typedef struct PgStat_MsgVacuum
-{
-       PgStat_MsgHdr m_hdr;
-       Oid                     m_databaseid;
-       Oid                     m_tableoid;
-       bool            m_analyze;
-       PgStat_Counter m_tuples;
-} PgStat_MsgVacuum;
+#define PGSTAT_MSG_PAYLOAD     (1000 - sizeof(PgStat_MsgHdr))
+
 
 /* ----------
- * PgStat_MsgAnalyze                   Sent by the backend or autovacuum daemon
- *                                                             after ANALYZE
+ * PgStat_MsgDummy                             A dummy message, ignored by the collector
  * ----------
  */
-typedef struct PgStat_MsgAnalyze
+typedef struct PgStat_MsgDummy
 {
        PgStat_MsgHdr m_hdr;
-       Oid                     m_databaseid;
-       Oid                     m_tableoid;
-       PgStat_Counter m_live_tuples;
-       PgStat_Counter m_dead_tuples;
-} PgStat_MsgAnalyze;
+} PgStat_MsgDummy;
 
 
 /* ----------
- * PgStat_MsgActivity                  Sent by the backends when they start
- *                                                             to parse a query.
+ * PgStat_TableEntry                   Per-table info in a MsgTabstat
  * ----------
  */
-#define PGSTAT_ACTIVITY_SIZE   PGSTAT_MSG_PAYLOAD
-
-typedef struct PgStat_MsgActivity
+typedef struct PgStat_TableEntry
 {
-       PgStat_MsgHdr m_hdr;
-       char            m_cmd_str[PGSTAT_ACTIVITY_SIZE];
-} PgStat_MsgActivity;
+       Oid                     t_id;
+       PgStat_TableCounts t_counts;
+} PgStat_TableEntry;
 
 /* ----------
  * PgStat_MsgTabstat                   Sent by the backend to report table
@@ -201,6 +189,7 @@ typedef struct PgStat_MsgTabstat
        PgStat_TableEntry m_entry[PGSTAT_NUM_TABENTRIES];
 } PgStat_MsgTabstat;
 
+
 /* ----------
  * PgStat_MsgTabpurge                  Sent by the backend to tell the collector
  *                                                             about dead tables.
@@ -243,6 +232,71 @@ typedef struct PgStat_MsgResetcounter
 } PgStat_MsgResetcounter;
 
 
+/* ----------
+ * PgStat_MsgAutovacStart              Sent by the autovacuum daemon to signal
+ *                                                             that a database is going to be processed
+ * ----------
+ */
+typedef struct PgStat_MsgAutovacStart
+{
+       PgStat_MsgHdr m_hdr;
+       Oid                     m_databaseid;
+       TimestampTz m_start_time;
+} PgStat_MsgAutovacStart;
+
+
+/* ----------
+ * PgStat_MsgVacuum                            Sent by the backend or autovacuum daemon
+ *                                                             after VACUUM or VACUUM ANALYZE
+ * ----------
+ */
+typedef struct PgStat_MsgVacuum
+{
+       PgStat_MsgHdr m_hdr;
+       Oid                     m_databaseid;
+       Oid                     m_tableoid;
+       bool            m_analyze;
+       bool            m_autovacuum;
+       TimestampTz m_vacuumtime;
+       PgStat_Counter m_tuples;
+} PgStat_MsgVacuum;
+
+
+/* ----------
+ * PgStat_MsgAnalyze                   Sent by the backend or autovacuum daemon
+ *                                                             after ANALYZE
+ * ----------
+ */
+typedef struct PgStat_MsgAnalyze
+{
+       PgStat_MsgHdr m_hdr;
+       Oid                     m_databaseid;
+       Oid                     m_tableoid;
+       bool            m_autovacuum;
+       TimestampTz m_analyzetime;
+       PgStat_Counter m_live_tuples;
+       PgStat_Counter m_dead_tuples;
+} PgStat_MsgAnalyze;
+
+
+/* ----------
+ * PgStat_MsgBgWriter           Sent by the bgwriter to update statistics.
+ * ----------
+ */
+typedef struct PgStat_MsgBgWriter
+{
+       PgStat_MsgHdr m_hdr;
+
+       PgStat_Counter  m_timed_checkpoints;
+       PgStat_Counter  m_requested_checkpoints;
+       PgStat_Counter  m_buf_written_checkpoints;
+       PgStat_Counter  m_buf_written_lru;
+       PgStat_Counter  m_buf_written_all;
+       PgStat_Counter  m_maxwritten_lru;
+       PgStat_Counter  m_maxwritten_all;
+} PgStat_MsgBgWriter;
+
+
 /* ----------
  * PgStat_Msg                                  Union over all possible messages.
  * ----------
@@ -251,8 +305,6 @@ typedef union PgStat_Msg
 {
        PgStat_MsgHdr msg_hdr;
        PgStat_MsgDummy msg_dummy;
-       PgStat_MsgBestart msg_bestart;
-       PgStat_MsgActivity msg_activity;
        PgStat_MsgTabstat msg_tabstat;
        PgStat_MsgTabpurge msg_tabpurge;
        PgStat_MsgDropdb msg_dropdb;
@@ -260,6 +312,7 @@ typedef union PgStat_Msg
        PgStat_MsgAutovacStart msg_autovacuum;
        PgStat_MsgVacuum msg_vacuum;
        PgStat_MsgAnalyze msg_analyze;
+       PgStat_MsgBgWriter msg_bgwriter;
 } PgStat_Msg;
 
 
@@ -271,7 +324,7 @@ typedef union PgStat_Msg
  * ------------------------------------------------------------
  */
 
-#define PGSTAT_FILE_FORMAT_ID  0x01A5BC94
+#define PGSTAT_FILE_FORMAT_ID  0x01A5BC96
 
 /* ----------
  * PgStat_StatDBEntry                  The collector's data per database
@@ -280,12 +333,15 @@ typedef union PgStat_Msg
 typedef struct PgStat_StatDBEntry
 {
        Oid                     databaseid;
-       int                     destroy;
-       int                     n_backends;
        PgStat_Counter n_xact_commit;
        PgStat_Counter n_xact_rollback;
        PgStat_Counter n_blocks_fetched;
        PgStat_Counter n_blocks_hit;
+       PgStat_Counter n_tuples_returned;
+       PgStat_Counter n_tuples_fetched;
+       PgStat_Counter n_tuples_inserted;
+       PgStat_Counter n_tuples_updated;
+       PgStat_Counter n_tuples_deleted;
        TimestampTz last_autovac_time;
 
        /*
@@ -296,53 +352,6 @@ typedef struct PgStat_StatDBEntry
 } PgStat_StatDBEntry;
 
 
-/* ----------
- * PgStat_StatBeEntry                  The collector's data per backend
- * ----------
- */
-typedef struct PgStat_StatBeEntry
-{
-       /* An entry is non-empty iff procpid > 0 */
-       int                     procpid;
-       TimestampTz start_timestamp;
-       TimestampTz activity_start_timestamp;
-
-       /*
-        * These fields are initialized by the BESTART message. If we have
-        * received messages from a backend before we have received its BESTART,
-        * these fields will be uninitialized: userid and databaseid will be
-        * InvalidOid, and clientaddr will be undefined.
-        */
-       Oid                     userid;
-       Oid                     databaseid;
-       SockAddr        clientaddr;
-
-       /*
-        * activity[] must be last in the struct, because we only write as much
-        * of it as needed to the stats file.
-        */
-       char            activity[PGSTAT_ACTIVITY_SIZE];
-} PgStat_StatBeEntry;
-
-
-/* ----------
- * PgStat_StatBeDead                   Because UDP packets can arrive out of
- *                                                             order, we need to keep some information
- *                                                             about backends that are known to be
- *                                                             dead for some seconds. This info is held
- *                                                             in a hash table of these structs.
- *
- * (This struct is not used in the stats file.)
- * ----------
- */
-typedef struct PgStat_StatBeDead
-{
-       int                     procpid;
-       int                     backendid;
-       int                     destroy;
-} PgStat_StatBeDead;
-
-
 /* ----------
  * PgStat_StatTabEntry                 The collector's data per table (or index)
  * ----------
@@ -350,7 +359,6 @@ typedef struct PgStat_StatBeDead
 typedef struct PgStat_StatTabEntry
 {
        Oid                     tableid;
-       int                     destroy;
 
        PgStat_Counter numscans;
 
@@ -367,31 +375,107 @@ typedef struct PgStat_StatTabEntry
 
        PgStat_Counter blocks_fetched;
        PgStat_Counter blocks_hit;
+
+       TimestampTz vacuum_timestamp;           /* user initiated vacuum */
+       TimestampTz autovac_vacuum_timestamp;           /* autovacuum initiated */
+       TimestampTz analyze_timestamp;          /* user initiated */
+       TimestampTz autovac_analyze_timestamp;          /* autovacuum initiated */
 } PgStat_StatTabEntry;
 
 
+/*
+ * Global statistics kept in the stats collector
+ */
+typedef struct PgStat_GlobalStats
+{
+       PgStat_Counter  timed_checkpoints;
+       PgStat_Counter  requested_checkpoints;
+       PgStat_Counter  buf_written_checkpoints;
+       PgStat_Counter  buf_written_lru;
+       PgStat_Counter  buf_written_all;
+       PgStat_Counter  maxwritten_lru;
+       PgStat_Counter  maxwritten_all;
+} PgStat_GlobalStats;
+
+
+/* ----------
+ * Shared-memory data structures
+ * ----------
+ */
+
+/* Max length of st_activity string ... perhaps replace with a GUC var? */
+#define PGBE_ACTIVITY_SIZE     1024
+
+/* ----------
+ * PgBackendStatus
+ *
+ * Each live backend maintains a PgBackendStatus struct in shared memory
+ * showing its current activity.  (The structs are allocated according to
+ * BackendId, but that is not critical.)  Note that the collector process
+ * has no involvement in, or even access to, these structs.
+ * ----------
+ */
+typedef struct PgBackendStatus
+{
+       /*
+        * To avoid locking overhead, we use the following protocol: a backend
+        * increments st_changecount before modifying its entry, and again after
+        * finishing a modification.  A would-be reader should note the value of
+        * st_changecount, copy the entry into private memory, then check
+        * st_changecount again.  If the value hasn't changed, and if it's even,
+        * the copy is valid; otherwise start over.  This makes updates cheap
+        * while reads are potentially expensive, but that's the tradeoff we want.
+        */
+       int                     st_changecount;
+
+       /* The entry is valid iff st_procpid > 0, unused if st_procpid == 0 */
+       int                     st_procpid;
+
+       /* Times when current backend, transaction, and activity started */
+       TimestampTz st_proc_start_timestamp;
+       TimestampTz st_txn_start_timestamp;
+       TimestampTz st_activity_start_timestamp;
+
+       /* Database OID, owning user's OID, connection client address */
+       Oid                     st_databaseid;
+       Oid                     st_userid;
+       SockAddr        st_clientaddr;
+
+       /* Is backend currently waiting on an lmgr lock? */
+       bool            st_waiting;
+
+       /* current command string; MUST be null-terminated */
+       char            st_activity[PGBE_ACTIVITY_SIZE];
+} PgBackendStatus;
+
+
 /* ----------
  * GUC parameters
  * ----------
  */
 extern bool pgstat_collect_startcollector;
 extern bool pgstat_collect_resetonpmstart;
-extern bool pgstat_collect_querystring;
 extern bool pgstat_collect_tuplelevel;
 extern bool pgstat_collect_blocklevel;
+extern bool pgstat_collect_querystring;
 
+/*
+ * BgWriter statistics counters are updated directly by bgwriter and bufmgr
+ */
+extern PgStat_MsgBgWriter BgWriterStats;
 
 /* ----------
  * Functions called from postmaster
  * ----------
  */
+extern Size BackendStatusShmemSize(void);
+extern void CreateSharedBackendStatus(void);
+
 extern void pgstat_init(void);
 extern int     pgstat_start(void);
-extern void pgstat_beterm(int pid);
 extern void pgstat_reset_all(void);
-
+extern void allow_immediate_pgstat_restart(void);
 #ifdef EXEC_BACKEND
-extern void PgstatBufferMain(int argc, char *argv[]);
 extern void PgstatCollectorMain(int argc, char *argv[]);
 #endif
 
@@ -400,99 +484,90 @@ extern void PgstatCollectorMain(int argc, char *argv[]);
  * Functions called from backends
  * ----------
  */
-extern void pgstat_bestart(void);
-
 extern void pgstat_ping(void);
-extern void pgstat_report_activity(const char *what);
-extern void pgstat_report_tabstat(void);
+
+extern void pgstat_report_tabstat(bool force);
+extern void pgstat_vacuum_tabstat(void);
+extern void pgstat_drop_database(Oid databaseid);
+extern void pgstat_drop_relation(Oid relid);
+
+extern void pgstat_clear_snapshot(void);
+extern void pgstat_reset_counters(void);
+
 extern void pgstat_report_autovac(Oid dboid);
 extern void pgstat_report_vacuum(Oid tableoid, bool shared,
                                         bool analyze, PgStat_Counter tuples);
 extern void pgstat_report_analyze(Oid tableoid, bool shared,
                                          PgStat_Counter livetuples,
                                          PgStat_Counter deadtuples);
-extern void pgstat_vacuum_tabstat(void);
-extern void pgstat_drop_relation(Oid relid);
 
-extern void pgstat_reset_counters(void);
+extern void pgstat_bestart(void);
+extern void pgstat_report_activity(const char *what);
+extern void pgstat_report_txn_timestamp(TimestampTz tstamp);
+extern void pgstat_report_waiting(bool waiting);
 
-extern void pgstat_initstats(PgStat_Info *stats, Relation rel);
+extern void pgstat_initstats(Relation rel);
 
+/* nontransactional event counts are simple enough to inline */
 
-#define pgstat_count_heap_scan(s)                                                                              \
+#define pgstat_count_heap_scan(rel)                                                                            \
        do {                                                                                                                            \
-               if (pgstat_collect_tuplelevel && (s)->tabentry != NULL)                 \
-                       ((PgStat_TableEntry *)((s)->tabentry))->t_numscans++;           \
+               if (pgstat_collect_tuplelevel && (rel)->pgstat_info != NULL)    \
+                       (rel)->pgstat_info->t_counts.t_numscans++;                                      \
        } while (0)
 /* kluge for bitmap scans: */
-#define pgstat_discount_heap_scan(s)                                                                   \
+#define pgstat_discount_heap_scan(rel)                                                                 \
        do {                                                                                                                            \
-               if (pgstat_collect_tuplelevel && (s)->tabentry != NULL)                 \
-                       ((PgStat_TableEntry *)((s)->tabentry))->t_numscans--;           \
+               if (pgstat_collect_tuplelevel && (rel)->pgstat_info != NULL)    \
+                       (rel)->pgstat_info->t_counts.t_numscans--;                                      \
        } while (0)
-#define pgstat_count_heap_getnext(s)                                                                   \
+#define pgstat_count_heap_getnext(rel)                                                                 \
        do {                                                                                                                            \
-               if (pgstat_collect_tuplelevel && (s)->tabentry != NULL)                 \
-                       ((PgStat_TableEntry *)((s)->tabentry))->t_tuples_returned++; \
+               if (pgstat_collect_tuplelevel && (rel)->pgstat_info != NULL)    \
+                       (rel)->pgstat_info->t_counts.t_tuples_returned++;                       \
        } while (0)
-#define pgstat_count_heap_fetch(s)                                                                             \
+#define pgstat_count_heap_fetch(rel)                                                                   \
        do {                                                                                                                            \
-               if (pgstat_collect_tuplelevel && (s)->tabentry != NULL)                 \
-                       ((PgStat_TableEntry *)((s)->tabentry))->t_tuples_fetched++; \
+               if (pgstat_collect_tuplelevel && (rel)->pgstat_info != NULL)    \
+                       (rel)->pgstat_info->t_counts.t_tuples_fetched++;                        \
        } while (0)
-#define pgstat_count_heap_insert(s)                                                                            \
+#define pgstat_count_index_scan(rel)                                                                   \
        do {                                                                                                                            \
-               if (pgstat_collect_tuplelevel && (s)->tabentry != NULL)                 \
-                       ((PgStat_TableEntry *)((s)->tabentry))->t_tuples_inserted++; \
+               if (pgstat_collect_tuplelevel && (rel)->pgstat_info != NULL)    \
+                       (rel)->pgstat_info->t_counts.t_numscans++;                                      \
        } while (0)
-#define pgstat_count_heap_update(s)                                                                            \
+#define pgstat_count_index_tuples(rel, n)                                                              \
        do {                                                                                                                            \
-               if (pgstat_collect_tuplelevel && (s)->tabentry != NULL)                 \
-                       ((PgStat_TableEntry *)((s)->tabentry))->t_tuples_updated++; \
+               if (pgstat_collect_tuplelevel && (rel)->pgstat_info != NULL)    \
+                       (rel)->pgstat_info->t_counts.t_tuples_returned += (n);          \
        } while (0)
-#define pgstat_count_heap_delete(s)                                                                            \
+#define pgstat_count_buffer_read(rel)                                                                  \
        do {                                                                                                                            \
-               if (pgstat_collect_tuplelevel && (s)->tabentry != NULL)                 \
-                       ((PgStat_TableEntry *)((s)->tabentry))->t_tuples_deleted++; \
+               if (pgstat_collect_blocklevel && (rel)->pgstat_info != NULL)    \
+                       (rel)->pgstat_info->t_counts.t_blocks_fetched++;                        \
        } while (0)
-#define pgstat_count_index_scan(s)                                                                             \
+#define pgstat_count_buffer_hit(rel)                                                                   \
        do {                                                                                                                            \
-               if (pgstat_collect_tuplelevel && (s)->tabentry != NULL)                 \
-                       ((PgStat_TableEntry *)((s)->tabentry))->t_numscans++;           \
-       } while (0)
-#define pgstat_count_index_tuples(s, n)                                                                        \
-       do {                                                                                                                            \
-               if (pgstat_collect_tuplelevel && (s)->tabentry != NULL)                 \
-                       ((PgStat_TableEntry *)((s)->tabentry))->t_tuples_returned += (n); \
-       } while (0)
-#define pgstat_count_buffer_read(s,r)                                                                  \
-       do {                                                                                                                            \
-               if (pgstat_collect_blocklevel) {                                                                \
-                       if ((s)->tabentry != NULL)                                                                      \
-                               ((PgStat_TableEntry *)((s)->tabentry))->t_blocks_fetched++; \
-                       else {                                                                                                          \
-                               pgstat_initstats((s), (r));                                                             \
-                               if ((s)->tabentry != NULL)                                                              \
-                                       ((PgStat_TableEntry *)((s)->tabentry))->t_blocks_fetched++; \
-                       }                                                                                                                       \
-               }                                                                                                                               \
-       } while (0)
-#define pgstat_count_buffer_hit(s,r)                                                                   \
-       do {                                                                                                                            \
-               if (pgstat_collect_blocklevel) {                                                                \
-                       if ((s)->tabentry != NULL)                                                                      \
-                               ((PgStat_TableEntry *)((s)->tabentry))->t_blocks_hit++; \
-                       else {                                                                                                          \
-                               pgstat_initstats((s), (r));                                                             \
-                               if ((s)->tabentry != NULL)                                                              \
-                                       ((PgStat_TableEntry *)((s)->tabentry))->t_blocks_hit++; \
-                       }                                                                                                                       \
-               }                                                                                                                               \
+               if (pgstat_collect_blocklevel && (rel)->pgstat_info != NULL)    \
+                       (rel)->pgstat_info->t_counts.t_blocks_hit++;                            \
        } while (0)
 
+extern void pgstat_count_heap_insert(Relation rel);
+extern void pgstat_count_heap_update(Relation rel);
+extern void pgstat_count_heap_delete(Relation rel);
+
+extern void AtEOXact_PgStat(bool isCommit);
+extern void AtEOSubXact_PgStat(bool isCommit, int nestDepth);
+
+extern void AtPrepare_PgStat(void);
+extern void PostPrepare_PgStat(void);
+
+extern void pgstat_twophase_postcommit(TransactionId xid, uint16 info,
+                                                                          void *recdata, uint32 len);
+extern void pgstat_twophase_postabort(TransactionId xid, uint16 info,
+                                                                         void *recdata, uint32 len);
 
-extern void pgstat_count_xact_commit(void);
-extern void pgstat_count_xact_rollback(void);
+extern void pgstat_send_bgwriter(void);
 
 /* ----------
  * Support functions for the SQL-callable functions to
@@ -501,7 +576,8 @@ extern void pgstat_count_xact_rollback(void);
  */
 extern PgStat_StatDBEntry *pgstat_fetch_stat_dbentry(Oid dbid);
 extern PgStat_StatTabEntry *pgstat_fetch_stat_tabentry(Oid relid);
-extern PgStat_StatBeEntry *pgstat_fetch_stat_beentry(int beid);
+extern PgBackendStatus *pgstat_fetch_stat_beentry(int beid);
 extern int     pgstat_fetch_stat_numbackends(void);
+extern PgStat_GlobalStats *pgstat_fetch_global(void);
 
 #endif   /* PGSTAT_H */