Parse and show hash buckets info

src/enums.ts

@@ -106,6 +106,8 @@ export enum Property {
   GROUP_COUNT = "Group Count",
   GROUP_KEY = "Group Key",
   HASH_CONDITION = "Hash Cond",
+  HASH_BATCHES = "Hash Batches",
+  HASH_BUCKETS = "Hash Buckets",
   HEAP_FETCHES = "Heap Fetches",
   INDEX_NAME = "Index Name",
   IO_READ_TIME = "I/O Read Time",
@@ -123,10 +125,13 @@ export enum Property {
   NODE_ID = "nodeId",
   NODE_TYPE = "Node Type",
   OPERATION = "Operation",
+  ORIGINAL_HASH_BATCHES = "Original Hash Batches",
+  ORIGINAL_HASH_BUCKETS = "Original Hash Buckets",
   OUTPUT = "Output",
   PARALLEL_AWARE = "Parallel Aware",
   PARENT_RELATIONSHIP = "Parent Relationship",
   PARTIAL_MODE = "Partial Mode",
+  PEAK_MEMORY_USAGE = "Peak Memory Usage",
   PEAK_SORT_SPACE_USED = "Peak Sort Space Used",
   PEV_PLAN_TAG = "plan_",
   PLANNER_ESTIMATE_DIRECTION = "*Planner Row Estimate Direction",

src/filters.ts

@@ -67,6 +67,13 @@ export function formatCost(value: unknown): string {
   return (value as number).toLocaleString(undefined, { minimumFractionDigits: 2 })
 }
 
+export function formatNumber(value: unknown): string {
+  if (value === undefined) {
+    return "N/A"
+  }
+  return (value as number).toLocaleString()
+}
+
 export function formatRows(value: unknown): string {
   if (value === undefined) {
     return "N/A"
@@ -245,6 +252,8 @@ const nodePropFormatters: Partial<Record<Property, Formatter>> = {
   [Property.EXCLUSIVE_TEMP_WRITTEN_BLOCKS]: formatBlocksHtml,
   [Property.FULL_SORT_GROUPS]: formatSortGroups,
   [Property.HEAP_FETCHES]: formatRows,
+  [Property.HASH_BATCHES]: formatNumber,
+  [Property.HASH_BUCKETS]: formatNumber,
   [Property.IO_READ_TIME]: formatDuration,
   [Property.IO_WRITE_TIME]: formatDuration,
   [Property.JIT]: formatJit,
@@ -254,8 +263,11 @@ const nodePropFormatters: Partial<Record<Property, Formatter>> = {
   [Property.LOCAL_IO_WRITE_TIME]: formatDuration,
   [Property.LOCAL_READ_BLOCKS]: formatBlocksHtml,
   [Property.LOCAL_WRITTEN_BLOCKS]: formatBlocksHtml,
+  [Property.ORIGINAL_HASH_BATCHES]: formatNumber,
+  [Property.ORIGINAL_HASH_BUCKETS]: formatNumber,
   [Property.OUTPUT]: formatList,
   [Property.PARALLEL_AWARE]: formatBoolean,
+  [Property.PEAK_MEMORY_USAGE]: formatKilobytes,
   [Property.PLANNER_ESTIMATE_DIRECTION]: formatEstimateDirection,
   [Property.PLANNER_ESTIMATE_FACTOR]: formatFactor,
   [Property.PLAN_ROWS]: formatRows,

src/services/__tests__/12-line-wrapped-plans/01-expect

@@ -54,7 +54,7 @@
             "Actual Total Time": 0.062,
             "Actual Rows": 2,
             "Actual Loops": 1,
-            "Buckets": 1024,
+            "Hash Buckets": 1024,
             "Plans": [
               {
                 "Node Type": "Seq Scan",

src/services/__tests__/12-line-wrapped-plans/02-expect

@@ -54,7 +54,7 @@
             "Actual Total Time": 0.062,
             "Actual Rows": 2,
             "Actual Loops": 1,
-            "Buckets": 1024,
+            "Hash Buckets": 1024,
             "Plans": [
               {
                 "Node Type": "Seq Scan",

src/services/__tests__/12-line-wrapped-plans/05-expect

@@ -111,7 +111,7 @@
             "Actual Total Time": 4.997,
             "Actual Rows": 10,
             "Actual Loops": 3,
-            "Buckets": 1024,
+            "Hash Buckets": 1024,
             "Shared Hit Blocks": 26,
             "Shared Read Blocks": 9,
             "Shared Written Blocks": 0,

src/services/__tests__/12-line-wrapped-plans/06-expect

@@ -127,7 +127,7 @@
                 "Actual Total Time": 0.555,
                 "Actual Rows": 10,
                 "Actual Loops": 3,
-                "Buckets": 1024,
+                "Hash Buckets": 1024,
                 "Shared Hit Blocks": 33,
                 "Shared Read Blocks": 0,
                 "Shared Written Blocks": 0,
@@ -164,7 +164,7 @@
             "Actual Total Time": 1.289,
             "Actual Rows": 10,
             "Actual Loops": 3,
-            "Buckets": 1024,
+            "Hash Buckets": 1024,
             "Shared Hit Blocks": 14,
             "Shared Read Blocks": 0,
             "Shared Written Blocks": 0,

src/services/__tests__/12-line-wrapped-plans/07-expect

@@ -127,7 +127,7 @@
                 "Actual Total Time": 0.4,
                 "Actual Rows": 10,
                 "Actual Loops": 3,
-                "Buckets": 1024,
+                "Hash Buckets": 1024,
                 "Shared Hit Blocks": 33,
                 "Shared Read Blocks": 0,
                 "Shared Written Blocks": 0,
@@ -164,7 +164,7 @@
             "Actual Total Time": 0.933,
             "Actual Rows": 10,
             "Actual Loops": 3,
-            "Buckets": 1024,
+            "Hash Buckets": 1024,
             "Shared Hit Blocks": 14,
             "Shared Read Blocks": 0,
             "Shared Written Blocks": 0,

src/services/__tests__/from-text/45-buckets-expect

@@ -0,0 +1,23 @@
+{
+  "Plan": {
+    "Node Type": "Hash Join",
+    "Plans": [
+      {
+        "Node Type": "Seq Scan"
+      }, {
+        "Node Type": "Hash",
+        "Shared Hit Blocks": 14909,
+        "Shared Read Blocks": 98728,
+        "Temp Written Blocks": 88109,
+        "Hash Buckets": 1048576,
+        "Hash Batches": 16,
+        "Peak Memory Usage": 62533,
+        "Plans": [
+          {
+            "Node Type": "Seq Scan"
+          }
+        ]
+      }
+    ]
+  }
+}

src/services/__tests__/from-text/45-buckets-plan

@@ -0,0 +1,23 @@
+                                                           QUERY PLAN                                                            
+---------------------------------------------------------------------------------------------------------------------------------
+ Hash Join  (cost=471037.90..1304640.12 rows=14999862 width=114) (actual time=1480.644..9895.695 rows=15001138 loops=1)
+   Hash Cond: (details.eid = entetes.eid)
+   Buffers: shared hit=15940 read=282899, temp read=246065 written=246065
+   I/O Timings: shared read=242.957, temp read=317.342 write=830.025
+   ->  Seq Scan on details  (cost=0.00..335215.62 rows=15001362 width=65) (actual time=0.014..1114.539 rows=15001289 loops=1)
+         Filter: (b IS NOT NULL)
+         Buffers: shared hit=1031 read=184171
+         I/O Timings: shared read=157.615
+   ->  Hash  (cost=238637.70..238637.70 rows=9999056 width=57) (actual time=1474.384..1474.385 rows=9999900 loops=1)
+         Buckets: 1048576  Batches: 16  Memory Usage: 62533kB
+         Buffers: shared hit=14909 read=98728, temp written=88109
+         I/O Timings: shared read=85.342, temp write=294.483
+         ->  Seq Scan on entetes  (cost=0.00..238637.70 rows=9999056 width=57) (actual time=0.016..567.370 rows=9999900 loops=1)
+               Filter: (a > 100)
+               Rows Removed by Filter: 101
+               Buffers: shared hit=14909 read=98728
+               I/O Timings: shared read=85.342
+ Planning:
+   Buffers: shared hit=8
+ Planning Time: 0.119 ms
+ Execution Time: 10244.597 ms

src/services/__tests__/from-text/46-buckets-expect

@@ -0,0 +1,24 @@
+{
+  "Plan": {
+    "Node Type": "Aggregate",
+    "Plans": [
+      {
+        "Node Type": "Hash Join",
+        "Plans": [
+          {
+            "Node Type": "Seq Scan"
+          },
+          {
+            "Node Type": "Hash",
+            "Actual Rows": 1001000,
+            "Hash Buckets": 262144,
+            "Original Hash Buckets": 262144,
+            "Hash Batches": 2048,
+            "Original Hash Batches": 8,
+            "Peak Memory Usage": 35157
+          }
+        ]
+      }
+    ]
+  }
+}

src/services/__tests__/from-text/46-buckets-plan

@@ -0,0 +1,56 @@
+-- source: https://www.postgresql.org/message-id/flat/CA%2BhUKGKWWmf%3DWELLG%3DaUGbcugRaSQbtm0tKYiBut-B2rVKX63g%40mail.gmail.com
+
+SET max_parallel_workers_per_gather = 0;
+SET synchronize_seqscans = off;
+SET work_mem = '4MB';
+
+
+
+CREATE TABLE r AS SELECT generate_series(1, 10000000)::int i;
+ANALYZE r;
+
+
+
+-- 1k uniform keys + 1m duplicates
+CREATE TABLE s1k (i int);
+INSERT INTO s1k SELECT generate_series(1, 1000)::int i;
+ALTER TABLE s1k SET (autovacuum_enabled = off);
+ANALYZE s1k;
+INSERT INTO s1k SELECT 42 FROM generate_series(1, 1000000);
+
+
+
+EXPLAIN ANALYZE SELECT COUNT(*) FROM r JOIN s1k USING (i);
+SET
+SET
+SET
+SELECT 10000000
+ANALYZE
+CREATE TABLE
+INSERT 0 1000
+ALTER TABLE
+ANALYZE
+INSERT 0 1000000
+                                                            QUERY PLAN                                                            
+----------------------------------------------------------------------------------------------------------------------------------
+ Aggregate  (cost=302291.30..302291.31 rows=1 width=8) (actual time=3656.686..3656.688 rows=1.00 loops=1)
+   Buffers: shared hit=6734 read=41984 dirtied=13621, temp read=49855 written=49855
+   ->  Hash Join  (cost=27826.00..300076.30 rows=886000 width=0) (actual time=146.635..3619.685 rows=1001000.00 loops=1)
+         Hash Cond: (r.i = s1k.i)
+         Buffers: shared hit=6734 read=41984 dirtied=13621, temp read=49855 written=49855
+         ->  Seq Scan on r  (cost=0.00..144294.76 rows=10000676 width=4) (actual time=0.201..1200.806 rows=10000000.00 loops=1)
+               Buffers: shared hit=2304 read=41984 dirtied=13621
+         ->  Hash  (cost=13290.00..13290.00 rows=886000 width=4) (actual time=145.449..145.450 rows=1001000.00 loops=1)
+               Buckets: 262144 (originally 262144)  Batches: 2048 (originally 8)  Memory Usage: 35157kB
+               Buffers: shared hit=4430, temp written=2930
+               ->  Seq Scan on s1k  (cost=0.00..13290.00 rows=886000 width=4) (actual time=0.011..61.783 rows=1001000.00 loops=1)
+                     Buffers: shared hit=4430
+ Planning:
+   Buffers: shared hit=83 read=9
+ Planning Time: 0.495 ms
+ JIT:
+   Functions: 10
+   Options: Inlining false, Optimization false, Expressions true, Deforming true
+   Timing: Generation 0.492 ms (Deform 0.134 ms), Inlining 0.000 ms, Optimization 0.266 ms, Emission 3.684 ms, Total 4.442 ms
+ Execution Time: 3674.128 ms
+(20 rows)

src/services/plan-service.ts

@@ -994,6 +994,10 @@ export class PlanService {
           return
         }
 
+        if (this.parseBuckets(extraMatches[2], element as Node)) {
+          return
+        }
+
         if (this.parseWAL(extraMatches[2], element as Node)) {
           return
         }
@@ -1120,6 +1124,30 @@ export class PlanService {
     el[_.map([type, method, "blocks"], _.capitalize).join(" ")] = value
   }
 
+  private parseBuckets(text: string, el: Node): boolean {
+
+    enum BucketsMatch {
+      Buckets = 1,
+      OriginalBuckets,
+      Batches,
+      OriginalBatches,
+      MemoryUsage,
+    }
+
+    const bucketsRegex = /Buckets:\s+([0-9]*)(?:\s*\(originally\s+([0-9]*)\))*\s+Batches:\s+([0-9]*)(?:\s*\(originally\s+([0-9]*)\))*\s+Memory Usage:\s+([0-9]*)kB.*\s*$/g
+    const bucketsMatches = bucketsRegex.exec(text)
+
+    if (bucketsMatches) {
+      el[Property.HASH_BUCKETS] = parseInt(bucketsMatches[BucketsMatch.Buckets], 0)
+      el[Property.HASH_BATCHES] = parseInt(bucketsMatches[BucketsMatch.Batches], 0)
+      el[Property.ORIGINAL_HASH_BUCKETS] = parseInt(bucketsMatches[BucketsMatch.OriginalBuckets], 0)
+      el[Property.ORIGINAL_HASH_BATCHES] = parseInt(bucketsMatches[BucketsMatch.OriginalBatches], 0)
+      el[Property.PEAK_MEMORY_USAGE] = parseInt(bucketsMatches[BucketsMatch.MemoryUsage], 0)
+      return true
+    }
+    return false
+  }
+
   private getWorker(node: Node, workerNumber: number): Worker | undefined {
     return _.find(node[Property.WORKERS], (worker) => {
       return worker[Property.WORKER_NUMBER] === workerNumber