Add num_done counter to the pg_stat_checkpointer view.

Checkpoints can be skipped when the server is idle. The existing num_timed and
num_requested counters in pg_stat_checkpointer track both completed and
skipped checkpoints, but there was no way to count only the completed ones.

This commit introduces the num_done counter, which tracks only completed
checkpoints, making it easier to see how many were actually performed.

Bump catalog version.

Author: Anton A. Melnikov
Reviewed-by: Fujii Masao
Discussion: https://postgr.es/m/[email protected]

Author: MasaoFujii

doc/src/sgml/monitoring.sgml

@@ -3063,7 +3063,16 @@ description | Waiting for a newly initialized WAL file to reach durable storage
        <structfield>num_requested</structfield> <type>bigint</type>
       </para>
       <para>
-       Number of requested checkpoints that have been performed
+       Number of backend requested checkpoints
+      </para></entry>
+     </row>
+
+      <row>
+      <entry role="catalog_table_entry"><para role="column_definition">
+       <structfield>num_done</structfield> <type>bigint</type>
+      </para>
+      <para>
+       Number of checkpoints that have been performed
       </para></entry>
      </row>
 

src/backend/access/transam/xlog.c

@@ -6878,8 +6878,11 @@ update_checkpoint_display(int flags, bool restartpoint, bool reset)
  * In this case, we only insert an XLOG_CHECKPOINT_SHUTDOWN record, and it's
  * both the record marking the completion of the checkpoint and the location
  * from which WAL replay would begin if needed.
+ *
+ * Returns true if a new checkpoint was performed, or false if it was skipped
+ * because the system was idle.
  */
-void
+bool
 CreateCheckPoint(int flags)
 {
 	bool		shutdown;
@@ -6971,7 +6974,7 @@ CreateCheckPoint(int flags)
 			END_CRIT_SECTION();
 			ereport(DEBUG1,
 					(errmsg_internal("checkpoint skipped because system is idle")));
-			return;
+			return false;
 		}
 	}
 
@@ -7353,6 +7356,8 @@ CreateCheckPoint(int flags)
 									 CheckpointStats.ckpt_segs_added,
 									 CheckpointStats.ckpt_segs_removed,
 									 CheckpointStats.ckpt_segs_recycled);
+
+	return true;
 }
 
 /*

src/backend/catalog/system_views.sql

@@ -1138,6 +1138,7 @@ CREATE VIEW pg_stat_checkpointer AS
     SELECT
         pg_stat_get_checkpointer_num_timed() AS num_timed,
         pg_stat_get_checkpointer_num_requested() AS num_requested,
+        pg_stat_get_checkpointer_num_performed() AS num_done,
         pg_stat_get_checkpointer_restartpoints_timed() AS restartpoints_timed,
         pg_stat_get_checkpointer_restartpoints_requested() AS restartpoints_req,
         pg_stat_get_checkpointer_restartpoints_performed() AS restartpoints_done,

src/backend/postmaster/checkpointer.c

@@ -460,10 +460,7 @@ CheckpointerMain(char *startup_data, size_t startup_data_len)
 			 * Do the checkpoint.
 			 */
 			if (!do_restartpoint)
-			{
-				CreateCheckPoint(flags);
-				ckpt_performed = true;
-			}
+				ckpt_performed = CreateCheckPoint(flags);
 			else
 				ckpt_performed = CreateRestartPoint(flags);
 
@@ -484,7 +481,7 @@ CheckpointerMain(char *startup_data, size_t startup_data_len)
 
 			ConditionVariableBroadcast(&CheckpointerShmem->done_cv);
 
-			if (ckpt_performed)
+			if (!do_restartpoint)
 			{
 				/*
 				 * Note we record the checkpoint start time not end time as
@@ -493,18 +490,32 @@ CheckpointerMain(char *startup_data, size_t startup_data_len)
 				 */
 				last_checkpoint_time = now;
 
-				if (do_restartpoint)
-					PendingCheckpointerStats.restartpoints_performed++;
+				if (ckpt_performed)
+					PendingCheckpointerStats.num_performed++;
 			}
 			else
 			{
-				/*
-				 * We were not able to perform the restartpoint (checkpoints
-				 * throw an ERROR in case of error).  Most likely because we
-				 * have not received any new checkpoint WAL records since the
-				 * last restartpoint. Try again in 15 s.
-				 */
-				last_checkpoint_time = now - CheckPointTimeout + 15;
+				if (ckpt_performed)
+				{
+					/*
+					 * The same as for checkpoint. Please see the
+					 * corresponding comment.
+					 */
+					last_checkpoint_time = now;
+
+					PendingCheckpointerStats.restartpoints_performed++;
+				}
+				else
+				{
+					/*
+					 * We were not able to perform the restartpoint
+					 * (checkpoints throw an ERROR in case of error).  Most
+					 * likely because we have not received any new checkpoint
+					 * WAL records since the last restartpoint. Try again in
+					 * 15 s.
+					 */
+					last_checkpoint_time = now - CheckPointTimeout + 15;
+				}
 			}
 
 			ckpt_active = false;

src/backend/utils/activity/pgstat_checkpointer.c

@@ -49,6 +49,7 @@ pgstat_report_checkpointer(void)
 #define CHECKPOINTER_ACC(fld) stats_shmem->stats.fld += PendingCheckpointerStats.fld
 	CHECKPOINTER_ACC(num_timed);
 	CHECKPOINTER_ACC(num_requested);
+	CHECKPOINTER_ACC(num_performed);
 	CHECKPOINTER_ACC(restartpoints_timed);
 	CHECKPOINTER_ACC(restartpoints_requested);
 	CHECKPOINTER_ACC(restartpoints_performed);
@@ -127,6 +128,7 @@ pgstat_checkpointer_snapshot_cb(void)
 #define CHECKPOINTER_COMP(fld) pgStatLocal.snapshot.checkpointer.fld -= reset.fld;
 	CHECKPOINTER_COMP(num_timed);
 	CHECKPOINTER_COMP(num_requested);
+	CHECKPOINTER_COMP(num_performed);
 	CHECKPOINTER_COMP(restartpoints_timed);
 	CHECKPOINTER_COMP(restartpoints_requested);
 	CHECKPOINTER_COMP(restartpoints_performed);

src/backend/utils/adt/pgstatfuncs.c

@@ -1191,6 +1191,12 @@ pg_stat_get_checkpointer_num_requested(PG_FUNCTION_ARGS)
 	PG_RETURN_INT64(pgstat_fetch_stat_checkpointer()->num_requested);
 }
 
+Datum
+pg_stat_get_checkpointer_num_performed(PG_FUNCTION_ARGS)
+{
+	PG_RETURN_INT64(pgstat_fetch_stat_checkpointer()->num_performed);
+}
+
 Datum
 pg_stat_get_checkpointer_restartpoints_timed(PG_FUNCTION_ARGS)
 {

src/include/access/xlog.h

@@ -239,7 +239,7 @@ extern void LocalProcessControlFile(bool reset);
 extern WalLevel GetActiveWalLevelOnStandby(void);
 extern void StartupXLOG(void);
 extern void ShutdownXLOG(int code, Datum arg);
-extern void CreateCheckPoint(int flags);
+extern bool CreateCheckPoint(int flags);
 extern bool CreateRestartPoint(int flags);
 extern WALAvailability GetWALAvailability(XLogRecPtr targetLSN);
 extern void XLogPutNextOid(Oid nextOid);

src/include/catalog/catversion.h

@@ -57,6 +57,6 @@
  */
 
 /*							yyyymmddN */
-#define CATALOG_VERSION_NO	202409271
+#define CATALOG_VERSION_NO	202409301
 
 #endif

src/include/catalog/pg_proc.dat

@@ -5820,6 +5820,11 @@
   proname => 'pg_stat_get_checkpointer_num_requested', provolatile => 's',
   proparallel => 'r', prorettype => 'int8', proargtypes => '',
   prosrc => 'pg_stat_get_checkpointer_num_requested' },
+{ oid => '8599',
+  descr => 'statistics: number of checkpoints performed by the checkpointer',
+  proname => 'pg_stat_get_checkpointer_num_performed', provolatile => 's',
+  proparallel => 'r', prorettype => 'int8', proargtypes => '',
+  prosrc => 'pg_stat_get_checkpointer_num_performed' },
 { oid => '6327',
   descr => 'statistics: number of timed restartpoints started by the checkpointer',
   proname => 'pg_stat_get_checkpointer_restartpoints_timed', provolatile => 's',

src/include/pgstat.h

@@ -294,6 +294,7 @@ typedef struct PgStat_CheckpointerStats
 {
 	PgStat_Counter num_timed;
 	PgStat_Counter num_requested;
+	PgStat_Counter num_performed;
 	PgStat_Counter restartpoints_timed;
 	PgStat_Counter restartpoints_requested;
 	PgStat_Counter restartpoints_performed;

src/test/regress/expected/rules.out

@@ -1824,6 +1824,7 @@ pg_stat_bgwriter| SELECT pg_stat_get_bgwriter_buf_written_clean() AS buffers_cle
     pg_stat_get_bgwriter_stat_reset_time() AS stats_reset;
 pg_stat_checkpointer| SELECT pg_stat_get_checkpointer_num_timed() AS num_timed,
     pg_stat_get_checkpointer_num_requested() AS num_requested,
+    pg_stat_get_checkpointer_num_performed() AS num_done,
     pg_stat_get_checkpointer_restartpoints_timed() AS restartpoints_timed,
     pg_stat_get_checkpointer_restartpoints_requested() AS restartpoints_req,
     pg_stat_get_checkpointer_restartpoints_performed() AS restartpoints_done,