Add "G" (server-side data generation) as an initialization step in pgbench.

This commit allows --init-steps option in pgbench to accept "G" character
meaning server-side data generation as an initialization step.
With "G", only limited queries are sent from pgbench client and
then data is actually generated in the server. This might make
the initialization phase faster if the bandwidth between pgbench client
and the server is low.

Author: Fabien Coelho
Reviewed-by: Anna Endo, Ibrar Ahmed, Fujii Masao
Discussion: https://postgr.es/m/alpine.DEB.2.21.1904061826420.3678@lancre

Author: MasaoFujii

doc/src/sgml/ref/pgbench.sgml

@@ -193,12 +193,34 @@ pgbench <optional> <replaceable>options</replaceable> </optional> <replaceable>d
           </listitem>
          </varlistentry>
          <varlistentry>
-          <term><literal>g</literal> (Generate data)</term>
+          <term><literal>g</literal> or <literal>G</literal> (Generate data, client-side or server-side)</term>
           <listitem>
            <para>
             Generate data and load it into the standard tables,
             replacing any data already present.
            </para>
+           <para>
+            With <literal>g</literal> (client-side data generation),
+            data is generated in <command>pgbench</command> client and then
+            sent to the server. This uses the client/server bandwidth
+            extensively through a <command>COPY</command>.
+            Using <literal>g</literal> causes logging to print one message
+            every 100,000 rows when generating data into
+            <structname>pgbench_accounts</structname> table.
+           </para>
+           <para>
+            With <literal>G</literal> (server-side data generation),
+            only limited queries are sent from <command>pgbench</command>
+            client and then data is actually generated in the server.
+            No significant bandwidth is required for this variant, but
+            the server will do more work.
+            Using <literal>G</literal> causes logging not to print any progress
+            message when generating data into
+            <structname>pgbench_accounts</structname> table.
+           </para>
+           <para>
+            
+           </para>
           </listitem>
          </varlistentry>
          <varlistentry>
@@ -262,9 +284,13 @@ pgbench <optional> <replaceable>options</replaceable> </optional> <replaceable>d
       <listitem>
        <para>
         Switch logging to quiet mode, producing only one progress message per 5
-        seconds. The default logging prints one message each 100000 rows, which
+        seconds. The default logging prints one message each 100,000 rows, which
         often outputs many lines per second (especially on good hardware).
        </para>
+       <para>
+        This setting has no effect if <literal>G</literal> is specified
+        in <option>-I</option>.
+       </para>
       </listitem>
      </varlistentry>
 

src/bin/pgbench/pgbench.c

@@ -132,6 +132,7 @@ static int	pthread_join(pthread_t th, void **thread_return);
  * some configurable parameters */
 
 #define DEFAULT_INIT_STEPS "dtgvp"	/* default -I setting */
+#define ALL_INIT_STEPS "dtgGvpf"	/* all possible steps */
 
 #define LOG_STEP_SECONDS	5	/* seconds between log messages */
 #define DEFAULT_NXACTS	10		/* default nxacts */
@@ -627,7 +628,7 @@ usage(void)
 		   "  %s [OPTION]... [DBNAME]\n"
 		   "\nInitialization options:\n"
 		   "  -i, --initialize         invokes initialization mode\n"
-		   "  -I, --init-steps=[dtgvpf]+ (default \"dtgvp\")\n"
+		   "  -I, --init-steps=[" ALL_INIT_STEPS "]+ (default \"" DEFAULT_INIT_STEPS "\")\n"
 		   "                           run selected initialization steps\n"
 		   "  -F, --fillfactor=NUM     set fill factor\n"
 		   "  -n, --no-vacuum          do not run VACUUM during initialization\n"
@@ -3803,10 +3804,23 @@ append_fillfactor(char *opts, int len)
 }
 
 /*
- * Fill the standard tables with some data
+ * Truncate away any old data, in one command in case there are foreign keys
  */
 static void
-initGenerateData(PGconn *con)
+initTruncateTables(PGconn *con)
+{
+	executeStatement(con, "truncate table "
+					 "pgbench_accounts, "
+					 "pgbench_branches, "
+					 "pgbench_history, "
+					 "pgbench_tellers");
+}
+
+/*
+ * Fill the standard tables with some data generated and sent from the client
+ */
+static void
+initGenerateDataClientSide(PGconn *con)
 {
 	char		sql[256];
 	PGresult   *res;
@@ -3820,23 +3834,16 @@ initGenerateData(PGconn *con)
 				remaining_sec;
 	int			log_interval = 1;
 
-	fprintf(stderr, "generating data...\n");
+	fprintf(stderr, "generating data (client-side)...\n");
 
 	/*
 	 * we do all of this in one transaction to enable the backend's
 	 * data-loading optimizations
 	 */
 	executeStatement(con, "begin");
 
-	/*
-	 * truncate away any old data, in one command in case there are foreign
-	 * keys
-	 */
-	executeStatement(con, "truncate table "
-					 "pgbench_accounts, "
-					 "pgbench_branches, "
-					 "pgbench_history, "
-					 "pgbench_tellers");
+	/* truncate away any old data */
+	initTruncateTables(con);
 
 	/*
 	 * fill branches, tellers, accounts in that order in case foreign keys
@@ -3940,6 +3947,51 @@ initGenerateData(PGconn *con)
 	executeStatement(con, "commit");
 }
 
+/*
+ * Fill the standard tables with some data generated on the server
+ *
+ * As already the case with the client-side data generation, the filler
+ * column defaults to NULL in pgbench_branches and pgbench_tellers,
+ * and is a blank-padded string in pgbench_accounts.
+ */
+static void
+initGenerateDataServerSide(PGconn *con)
+{
+	char		sql[256];
+
+	fprintf(stderr, "generating data (server-side)...\n");
+
+	/*
+	 * we do all of this in one transaction to enable the backend's
+	 * data-loading optimizations
+	 */
+	executeStatement(con, "begin");
+
+	/* truncate away any old data */
+	initTruncateTables(con);
+
+	snprintf(sql, sizeof(sql),
+			 "insert into pgbench_branches(bid,bbalance) "
+			 "select bid, 0 "
+			 "from generate_series(1, %d) as bid", nbranches * scale);
+	executeStatement(con, sql);
+
+	snprintf(sql, sizeof(sql),
+			 "insert into pgbench_tellers(tid,bid,tbalance) "
+			 "select tid, (tid - 1) / %d + 1, 0 "
+			 "from generate_series(1, %d) as tid", ntellers, ntellers * scale);
+	executeStatement(con, sql);
+
+	snprintf(sql, sizeof(sql),
+			 "insert into pgbench_accounts(aid,bid,abalance,filler) "
+			 "select aid, (aid - 1) / %d + 1, 0, '' "
+			 "from generate_series(1, "INT64_FORMAT") as aid",
+			 naccounts, (int64) naccounts * scale);
+	executeStatement(con, sql);
+
+	executeStatement(con, "commit");
+}
+
 /*
  * Invoke vacuum on the standard tables
  */
@@ -4020,21 +4072,21 @@ initCreateFKeys(PGconn *con)
 static void
 checkInitSteps(const char *initialize_steps)
 {
-	const char *step;
-
 	if (initialize_steps[0] == '\0')
 	{
 		fprintf(stderr, "no initialization steps specified\n");
 		exit(1);
 	}
 
-	for (step = initialize_steps; *step != '\0'; step++)
+	for (const char *step = initialize_steps; *step != '\0'; step++)
 	{
-		if (strchr("dtgvpf ", *step) == NULL)
+		if (strchr(ALL_INIT_STEPS " ", *step) == NULL)
 		{
-			fprintf(stderr, "unrecognized initialization step \"%c\"\n",
+			fprintf(stderr,
+					"unrecognized initialization step \"%c\"\n",
 					*step);
-			fprintf(stderr, "allowed steps are: \"d\", \"t\", \"g\", \"v\", \"p\", \"f\"\n");
+			fprintf(stderr,
+					"Allowed step characters are: \"" ALL_INIT_STEPS "\".\n");
 			exit(1);
 		}
 	}
@@ -4075,8 +4127,12 @@ runInitSteps(const char *initialize_steps)
 				initCreateTables(con);
 				break;
 			case 'g':
-				op = "generate";
-				initGenerateData(con);
+				op = "client-side generate";
+				initGenerateDataClientSide(con);
+				break;
+			case 'G':
+				op = "server-side generate";
+				initGenerateDataServerSide(con);
 				break;
 			case 'v':
 				op = "vacuum";

src/bin/pgbench/t/001_pgbench_with_server.pl

@@ -130,15 +130,15 @@ sub pgbench
 
 # Test interaction of --init-steps with legacy step-selection options
 pgbench(
-	'--initialize --init-steps=dtpvgvv --no-vacuum --foreign-keys --unlogged-tables --partitions=3',
+	'--initialize --init-steps=dtpvGvv --no-vacuum --foreign-keys --unlogged-tables --partitions=3',
 	0,
 	[qr{^$}],
 	[
 		qr{dropping old tables},
 		qr{creating tables},
 		qr{creating 3 partitions},
 		qr{creating primary keys},
-		qr{.* of .* tuples \(.*\) done},
+		qr{generating data \(server-side\)},
 		qr{creating foreign keys},
 		qr{(?!vacuuming)}, # no vacuum
 		qr{done in \d+\.\d\d s }

src/bin/pgbench/t/002_pgbench_no_server.pl

@@ -147,7 +147,7 @@ sub pgbench_scripts
 	[
 		'invalid init step',
 		'-i -I dta',
-		[ qr{unrecognized initialization step}, qr{allowed steps are} ]
+		[ qr{unrecognized initialization step}, qr{Allowed step characters are} ]
 	],
 	[
 		'bad random seed',