Merge cockroachdb#32704

32704: workload/ycsb: add flag to use column families r=nvanbenschoten a=nvanbenschoten

This change adds a new `--families` flag to the ycsb workload. Now
that cockroachdb#18168 is addressed, this significantly reduces the contention
present in the workload by avoiding conflicts on updates to different
columns in the same table.

I just confirmed that this still provides a huge speedup. On a 24 cpu machine:
```
workload run ycsb --init --workload='A' --concurrency=128 --duration=1m


--families=false gives:

_elapsed___errors_____ops(total)___ops/sec(cum)__avg(ms)__p50(ms)__p95(ms)__p99(ms)_pMax(ms)__total
   60.0s        0         103089         1718.0     13.9      0.6      1.8    604.0   1476.4  read
   60.0s        0         102947         1715.6     59.5      5.2     11.5   2281.7   8321.5  update

_elapsed___errors_____ops(total)___ops/sec(cum)__avg(ms)__p50(ms)__p95(ms)__p99(ms)_pMax(ms)__result
   60.0s        0         206036         3433.6     36.7      3.3      8.9   1342.2   8321.5


--families=true gives:

_elapsed___errors_____ops(total)___ops/sec(cum)__avg(ms)__p50(ms)__p95(ms)__p99(ms)_pMax(ms)__total
   60.0s        0         333477         5557.8      9.2      0.6      6.0    302.0   1275.1  read
   60.0s        0         332366         5539.3     13.7      6.8     17.8     54.5   4831.8  update

_elapsed___errors_____ops(total)___ops/sec(cum)__avg(ms)__p50(ms)__p95(ms)__p99(ms)_pMax(ms)__result
   60.0s        0         665843        11097.1     11.5      3.9     16.3    268.4   4831.8
```

cc. @robert-s-lee @drewdeally 

Release note: None

Co-authored-by: Nathan VanBenschoten <[email protected]>

pkg/workload/ycsb/ycsb.go

@@ -51,6 +51,30 @@ const (
 		FIELD8 TEXT,
 		FIELD9 TEXT
 	)`
+	usertableSchemaRelationalWithFamilies = `(
+		ycsb_key VARCHAR(255) PRIMARY KEY NOT NULL,
+		FIELD0 TEXT,
+		FIELD1 TEXT,
+		FIELD2 TEXT,
+		FIELD3 TEXT,
+		FIELD4 TEXT,
+		FIELD5 TEXT,
+		FIELD6 TEXT,
+		FIELD7 TEXT,
+		FIELD8 TEXT,
+		FIELD9 TEXT,
+		FAMILY (ycsb_key),
+		FAMILY (FIELD0),
+		FAMILY (FIELD1),
+		FAMILY (FIELD2),
+		FAMILY (FIELD3),
+		FAMILY (FIELD4),
+		FAMILY (FIELD5),
+		FAMILY (FIELD6),
+		FAMILY (FIELD7),
+		FAMILY (FIELD8),
+		FAMILY (FIELD9)
+	)`
 	usertableSchemaJSON = `(
 		ycsb_key VARCHAR(255) PRIMARY KEY NOT NULL,
 		FIELD JSONB
@@ -64,6 +88,7 @@ type ycsb struct {
 	seed        int64
 	initialRows int
 	json        bool
+	families    bool
 	splits      int
 
 	workload                                   string
@@ -89,6 +114,7 @@ var ycsbMeta = workload.Meta{
 		g.flags.IntVar(&g.initialRows, `initial-rows`, 10000,
 			`Initial number of rows to sequentially insert before beginning Zipfian workload`)
 		g.flags.BoolVar(&g.json, `json`, false, `Use JSONB rather than relational data`)
+		g.flags.BoolVar(&g.families, `families`, true, `Place each column in its own column family`)
 		g.flags.IntVar(&g.splits, `splits`, 0, `Number of splits to perform before starting normal operations`)
 		g.flags.StringVar(&g.workload, `workload`, `B`, `Workload type. Choose from A-F.`)
 		g.flags.StringVar(&g.distribution, `request-distribution`, `zipfian`, `Distribution for random number generator [zipfian, uniform].`)
@@ -169,7 +195,11 @@ func (g *ycsb) Tables() []workload.Table {
 	if g.json {
 		usertable.Schema = usertableSchemaJSON
 	} else {
-		usertable.Schema = usertableSchemaRelational
+		if g.families {
+			usertable.Schema = usertableSchemaRelationalWithFamilies
+		} else {
+			usertable.Schema = usertableSchemaRelational
+		}
 	}
 	return []workload.Table{usertable}
 }