Merge #150321

150321: **workload: add main orchestrator `workload.go` for end-to-end data g… r=nameisbhaskar a=Pradyum-Git

…eneration**

Previously, the workload generator package provided parsing of input DDL and metadata structs, as well as a set of pluggable generators and wrappers, but lacked a central entry point to drive the entire workflow.

This patch introduces a new `workload.go` file that serves as the main orchestrator:

* Defines the `main` function and CLI flags (e.g. `--ddl-path`, `--output-dir`, batch size, random seed).
* Invokes DDL parsing to build in-memory schema representations.
* Constructs per-table generators by calling into the existing factory and applying wrappers.
* Streams data generation for each table, writing SQL (or CSV) to the specified output.

Fixes: CRDB-51752

Release note (cli change): Adds `workload.go` as the primary CLI entry point for workload generation, wiring together DDL parsing, schema construction, generator factory, and output routines—establishing a cohesive, flag-driven workflow for end-to-end data generation.

Co-authored-by: Pradyum <[email protected]>

Author: craig[bot]

pkg/ccl/workloadccl/allccl/BUILD.bazel

@@ -32,6 +32,7 @@ go_library(
         "//pkg/workload/ttlbench",
         "//pkg/workload/ttllogger",
         "//pkg/workload/vecann",
+        "//pkg/workload/workload_generator",
         "//pkg/workload/ycsb",
     ],
 )

pkg/ccl/workloadccl/allccl/all.go

@@ -36,5 +36,6 @@ import (
 	_ "github.com/cockroachdb/cockroach/pkg/workload/ttlbench"
 	_ "github.com/cockroachdb/cockroach/pkg/workload/ttllogger"
 	_ "github.com/cockroachdb/cockroach/pkg/workload/vecann"
+	_ "github.com/cockroachdb/cockroach/pkg/workload/workload_generator"
 	_ "github.com/cockroachdb/cockroach/pkg/workload/ycsb"
 )

pkg/ccl/workloadccl/allccl/all_test.go

@@ -147,6 +147,9 @@ func TestAllRegisteredSetup(t *testing.T) {
 			// This requires downloading from a GCP bucket and storing in the
 			// machine's ~/.cache directory.
 			continue
+		case `workload_generator`:
+			// This will take its schema generation data from flags at run time, so static checks are not valid.
+			continue
 		}
 
 		t.Run(meta.Name, func(t *testing.T) {

pkg/cli/BUILD.bazel

@@ -244,6 +244,7 @@ go_library(
         "//pkg/workload/tpch",
         "//pkg/workload/ttlbench",
         "//pkg/workload/ttllogger",
+        "//pkg/workload/workload_generator",
         "//pkg/workload/workloadsql",
         "//pkg/workload/ycsb",
         "@com_github_cockroachdb_apd_v3//:apd",

pkg/cli/cli.go

@@ -30,17 +30,18 @@ import (
 	_ "github.com/cockroachdb/cockroach/pkg/workload/bank"       // registers workloads
 	_ "github.com/cockroachdb/cockroach/pkg/workload/bulkingest" // registers workloads
 	workloadcli "github.com/cockroachdb/cockroach/pkg/workload/cli"
-	_ "github.com/cockroachdb/cockroach/pkg/workload/debug"     // registers workloads
-	_ "github.com/cockroachdb/cockroach/pkg/workload/examples"  // registers workloads
-	_ "github.com/cockroachdb/cockroach/pkg/workload/insights"  // registers workloads
-	_ "github.com/cockroachdb/cockroach/pkg/workload/kv"        // registers workloads
-	_ "github.com/cockroachdb/cockroach/pkg/workload/movr"      // registers workloads
-	_ "github.com/cockroachdb/cockroach/pkg/workload/sqlstats"  // registers workloads
-	_ "github.com/cockroachdb/cockroach/pkg/workload/tpcc"      // registers workloads
-	_ "github.com/cockroachdb/cockroach/pkg/workload/tpch"      // registers workloads
-	_ "github.com/cockroachdb/cockroach/pkg/workload/ttlbench"  // registers workloads
-	_ "github.com/cockroachdb/cockroach/pkg/workload/ttllogger" // registers workloads
-	_ "github.com/cockroachdb/cockroach/pkg/workload/ycsb"      // registers workloads
+	_ "github.com/cockroachdb/cockroach/pkg/workload/debug"              // registers workloads
+	_ "github.com/cockroachdb/cockroach/pkg/workload/examples"           // registers workloads
+	_ "github.com/cockroachdb/cockroach/pkg/workload/insights"           // registers workloads
+	_ "github.com/cockroachdb/cockroach/pkg/workload/kv"                 // registers workloads
+	_ "github.com/cockroachdb/cockroach/pkg/workload/movr"               // registers workloads
+	_ "github.com/cockroachdb/cockroach/pkg/workload/sqlstats"           // registers workloads
+	_ "github.com/cockroachdb/cockroach/pkg/workload/tpcc"               // registers workloads
+	_ "github.com/cockroachdb/cockroach/pkg/workload/tpch"               // registers workloads
+	_ "github.com/cockroachdb/cockroach/pkg/workload/ttlbench"           // registers workloads
+	_ "github.com/cockroachdb/cockroach/pkg/workload/ttllogger"          // registers workloads
+	_ "github.com/cockroachdb/cockroach/pkg/workload/workload_generator" //registers workloads
+	_ "github.com/cockroachdb/cockroach/pkg/workload/ycsb"               // registers workloads
 	"github.com/cockroachdb/errors"
 	"github.com/spf13/cobra"
 )

pkg/workload/workload_generator/BUILD.bazel

@@ -9,12 +9,20 @@ go_library(
         "schema_generator.go",
         "types.go",
         "utils.go",
+        "workload.go",
     ],
     importpath = "github.com/cockroachdb/cockroach/pkg/workload/workload_generator",
     visibility = ["//visibility:public"],
     deps = [
+        "//pkg/col/coldata",
+        "//pkg/sql/types",
+        "//pkg/util/bufalloc",
         "//pkg/util/timeutil",
+        "//pkg/workload",
+        "//pkg/workload/histogram",
         "@com_github_cockroachdb_errors//:errors",
+        "@com_github_spf13_pflag//:pflag",
+        "@in_gopkg_yaml_v2//:yaml_v2",
     ],
 )