-
Notifications
You must be signed in to change notification settings - Fork 148
/
Copy pathspark-sql-joins.html
323 lines (287 loc) · 14.2 KB
/
spark-sql-joins.html
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
310
311
312
313
314
315
316
317
318
319
320
321
322
323
<!doctype html>
<html lang="en">
<head>
<meta charset="utf-8">
<meta name="viewport" content="width=device-width, initial-scale=1.0, maximum-scale=1.0, user-scalable=no">
<title>Apache Spark™ Workshop | Spark SQL | Joins</title>
<meta name="description" content="Apache Spark™ Workshop | Spark SQL | Joins">
<meta name="author" content="Jacek Laskowski">
<link rel="stylesheet" href="reveal.js/css/reveal.css">
<link rel="stylesheet" href="reveal.js/css/theme/beige.css">
<!-- Theme used for syntax highlighting of code -->
<link rel="stylesheet" href="reveal.js/lib/css/zenburn.css">
<!-- Jacek: custom formatting -->
<link rel="stylesheet" href="revealjs-css/jacek.css">
<!-- Printing and PDF exports -->
<script>
var link = document.createElement('link');
link.rel = 'stylesheet';
link.type = 'text/css';
link.href = window.location.search.match(/print-pdf/gi) ? 'reveal.js/css/print/pdf.css' : 'reveal.js/css/print/paper.css';
document.getElementsByTagName('head')[0].appendChild(link);
</script>
</head>
<body>
<div class="reveal">
<div class="footer">
<footer style="font-size: small;">
© <a href="https://medium.com/@jaceklaskowski">Jacek Laskowski</a> 2022 / <a href="https://twitter.com/jaceklaskowski">@JacekLaskowski</a>
</footer>
</div>
<div class="slides">
<section class="intro" data-transition="zoom" id="home">
<p>
<img width="17%" style="background:none; border:none; box-shadow:none;" data-src="images/spark-logo.png">
<img width="10%" src="images/jacek_laskowski_20201229_200x200.png" style="border: 0">
</p>
<h1>Joins</h1>
<br />
<h3>Apache Spark 3.2 / Spark SQL</h3>
<hr />
<h4 style="font-size: smaller;">
<a href="https://twitter.com/jaceklaskowski">@jaceklaskowski</a> / <a
href="https://stackoverflow.com/users/1305344/jacek-laskowski">StackOverflow</a> / <a
href="https://github.com/jaceklaskowski">GitHub</a> / <a
href="https://www.linkedin.com/in/jaceklaskowski/">LinkedIn</a>
<br>
The "Internals" Books: <a href="https://books.japila.pl/apache-spark-internals">Apache Spark</a> / <a
href="https://books.japila.pl/spark-sql-internals/">Spark SQL</a> / <a
href="https://books.japila.pl/delta-lake-internals">Delta Lake</a>
</h4>
</section>
<section id="agenda" data-markdown>
<textarea data-template>
## Agenda
1. [join Operators](#/join-dataset)
1. [Join Condition](#/join-condition)
1. [Join Types](#/join-types)
1. [Join Clause (SQL)](#/join-sql)
1. [Broadcast Join (aka Map-Side Join)](#/broadcast-join)
1. [Join Internals and Optimizations](#/join-internals)
</textarea>
</section>
<section id="join-dataset" data-markdown>
<textarea data-template>
## join Operators
1. You can join two Datasets using join operators:
* **join** for untyped Row-based joins
* Type-preserving **joinWith**
* **crossJoin** for explicit cartesian joins
1. Join conditions in **join** or **filter** / **where** operators
<pre><code class="scala" data-trim style="word-wrap: break-word;">
// equivalent to left("id") === right("id")
left.join(right, "id")
left.join(right, "id", "anti") // explicit join type
left.join(right).filter(left("id") === right("id"))
left.join(right).where(left("id") === right("id"))
</code></pre>
<!-- .element: style="width: 810px;" -->
1. Switch to [The Internals of Spark SQL](https://bit.ly/spark-sql-internals)
* [Dataset Join Operators](https://jaceklaskowski.gitbooks.io/mastering-spark-sql/content/spark-sql-joins.html)
</textarea>
</section>
<section id="join-condition" data-markdown>
<textarea data-template>
## Join Condition
1. Join conditions in **join** or **filter** / **where** operators
<pre><code class="scala" data-trim style="word-wrap: break-word;">
// equivalent to left("id") === right("id")
left.join(right, "id")
// explicit join type
left.join(right, "id", "anti")
left.join(right).filter(left("id") === right("id"))
left.join(right).where(left("id") === right("id"))
// expr AND expr in multiple where's
left.join(right)
.where(left("id") === right("empid"))
.where(left("name") === right("empname"))
</code></pre>
<!-- .element: style="width: 810px;" -->
1. Uses [Column API](http://spark.apache.org/docs/latest/api/scala/index.html#org.apache.spark.sql.Column)
```
left("id") === right("empid")
left("id") === right("empid") AND left("name") === right("empname")
```
<!-- .element: style="width: 850px;" -->
</textarea>
</section>
<section id="join-types" data-markdown>
<textarea data-template>
## Join Types
| SQL | Names |
| ------------|-------------|
| CROSS | cross |
| INNER | inner |
| FULL OUTER | outer, full, fullouter |
| LEFT ANTI | leftanti |
| LEFT OUTER | leftouter, left |
| LEFT SEMI | leftsemi |
| RIGHT OUTER | rightouter, right |
<!-- .element: style="font-size: 0.8em;" -->
</textarea>
</section>
<section id="join-sql" data-markdown>
<textarea data-template>
## Join Clause (SQL)
1. You can join two tables using regular SQL join operators
1. Use **SparkSession.sql** to execute SQL
```scala
spark.sql("select * from t1, t2 where t1.id = t2.id")
```
1. Switch to [The Internals of Spark SQL](https://bit.ly/spark-sql-internals)
* [Executing SQL Queries (aka SQL Mode) — sql Method](https://jaceklaskowski.gitbooks.io/mastering-spark-sql/content/spark-sql-SparkSession.html#sql)
</textarea>
</section>
<section id="broadcast-join" data-markdown>
<textarea data-template>
<!-- .slide: style="font-size: 90%" -->
## Broadcast Join (aka Map-Side Join)
1. Use **broadcast** function to mark a Dataset to be broadcast
```scala
left.join(broadcast(right), "token")
```
1. Use **broadcast hints** in SQL
```
SELECT /*+ BROADCAST (t1) */ * FROM t1, t2 WHERE t1.id = t2.id
```
* **BROADCAST**, **BROADCASTJOIN** or **MAPJOIN** hints supported
1. Switch to [The Internals of Spark SQL](https://bit.ly/spark-sql-internals)
* [broadcast Function](https://jaceklaskowski.gitbooks.io/mastering-spark-sql/content/spark-sql-functions.html#broadcast)
* [withHints Parsing Handler](https://jaceklaskowski.gitbooks.io/mastering-spark-sql/content/spark-sql-AstBuilder.html#withHints)
</textarea>
</section>
<section>
<section id="join-internals">
<h1>Join</h1>
<h1 style="font-size: 2.77em;">Internals and Optimizations</h1>
</section>
<section id="join-logical-operator">
<h2>Join Logical Operator</h2>
<ol>
<li><b>Join</b> is a binary logical operator with two logical operators, a join type and an optional join expression</li>
<li>Switch to <a href="https://bit.ly/spark-sql-internals">The Internals of Spark SQL</a>
<ul>
<li><a href="https://jaceklaskowski.gitbooks.io/mastering-spark-sql/content/spark-sql-LogicalPlan-Join.html">Join Logical Operator</a></li>
</ul>
</li>
</ol>
</section>
<section id="join-selection" style="font-size: 80%;">
<h2>JoinSelection Execution Planning Strategy</h2>
<ol>
<li><b>JoinSelection</b> is an execution planning strategy that SparkPlanner uses to plan Join logical operators</li>
<li>Supported joins (and their physical operators)
<ol>
<li><b>Broadcast Hash Join</b> (BroadcastHashJoinExec)</li>
<li><b>Shuffled Hash Join</b> (ShuffledHashJoinExec)</li>
<li><b>Sort Merge Join</b> (SortMergeJoinExec)</li>
<li><b>Broadcast Nested Loop Join</b> (BroadcastNestedLoopJoinExec)</li>
<li><b>Cartesian Join</b> (CartesianProductExec)</li>
</ol>
</li>
<li>Switch to <a href="https://bit.ly/spark-sql-internals">The Internals of Spark SQL</a>
<ul>
<li><a href="https://jaceklaskowski.gitbooks.io/mastering-spark-sql/content/spark-sql-SparkStrategy-JoinSelection.html">JoinSelection Execution Planning Strategy</a></li>
</ul>
</li>
</ol>
</section>
<section id="join-optimization-bucketing">
<h2>Join Optimization — Bucketing</h2>
<ol>
<li><b>Bucketing</b> is an optimization technique in Spark SQL that uses buckets and bucketing columns to determine data partitioning.</li>
<li>Optimize performance of a join query by reducing shuffles (aka exchanges)</li>
<li>Switch to <a href="https://bit.ly/spark-sql-internals">The Internals of Spark SQL</a>
<ul>
<li><a href="https://jaceklaskowski.gitbooks.io/mastering-spark-sql/content/spark-sql-bucketing.html">Bucketing</a></li>
</ul>
</li>
<li>My talk <a href="https://youtu.be/dv7IIYuQOXI">Bucketing in Spark SQL 2 3</a> on Spark+AI Summit 2018</li>
</ol>
</section>
<section id="join-optimization-join-reorg">
<h2>Join Optimization — Join Reordering</h2>
<ol>
<li><b>Join Reordering</b> is an optimization of a logical query plan that the Spark Optimizer uses for joins</li>
<li>Switch to <a href="https://bit.ly/spark-sql-internals">The Internals of Spark SQL</a>
<ul>
<li><a href="https://jaceklaskowski.gitbooks.io/mastering-spark-sql/content/spark-sql-Optimizer-ReorderJoin.html">ReorderJoin Logical Query Optimization</a></li>
</ul>
</li>
</ol>
</section>
<section id="join-optimization-join-reorg-cbo" style="font-size: 80%;">
<h2>Join Optimization — Cost-based Join Reordering</h2>
<ol>
<li><b>Cost-based Join Reordering</b> is an optimization of a logical query plan that the Spark Optimizer uses for joins in cost-based optimization</li>
<li>Switch to <a href="https://bit.ly/spark-sql-internals">The Internals of Spark SQL</a>
<ul>
<li><a href="https://jaceklaskowski.gitbooks.io/mastering-spark-sql/content/spark-sql-Optimizer-CostBasedJoinReorder.html">CostBasedJoinReorder Logical Query Optimization</a></li>
</ul>
</li>
</ol>
</section>
</section>
<section id="recap" data-markdown>
<textarea data-template>
## Recap
1. [join Operators](#/join-dataset)
1. [Join Condition](#/join-condition)
1. [Join Types](#/join-types)
1. [Join Clause (SQL)](#/join-sql)
1. [Broadcast Join (aka Map-Side Join)](#/broadcast-join)
1. [Join Internals and Optimizations](#/join-internals)
</textarea>
</section>
<section style="text-align: left" data-markdown id="questions">
<textarea data-template>
# Questions?
* Read [The Internals of Apache Spark](https://bit.ly/apache-spark-internals)
* Read [The Internals of Spark SQL](https://bit.ly/spark-sql-internals)
* Read [The Internals of Spark Structured Streaming](https://bit.ly/spark-structured-streaming)
* Follow [@jaceklaskowski](https://twitter.com/jaceklaskowski) on twitter
* Upvote [my questions and answers on StackOverflow](http://stackoverflow.com/users/1305344/jacek-laskowski)
</textarea>
</section>
</div>
</div>
<script src="reveal.js/lib/js/head.min.js"></script>
<script src="reveal.js/js/reveal.js"></script>
<script>
// More info about config & dependencies:
// - https://github.com/hakimel/reveal.js#configuration
// - https://github.com/hakimel/reveal.js#dependencies
Reveal.initialize({
controls: true,
progress: true,
history: true,
center: true,
slideNumber: true,
transition: 'slide', // none/fade/slide/convex/concave/zoom
menu: {
markers: true,
openSlideNumber: true
},
dependencies: [
{ src: 'reveal.js/lib/js/classList.js', condition: function () { return !document.body.classList; } },
{ src: 'reveal.js/plugin/markdown/marked.js' },
{ src: 'reveal.js/plugin/markdown/markdown.js' },
{ src: 'reveal.js/plugin/zoom-js/zoom.js', async: true },
{ src: 'reveal.js/plugin/notes/notes.js', async: true },
{ src: 'reveal.js/plugin/highlight/highlight.js', async: true, callback: function () { hljs.initHighlightingOnLoad(); } }
]
});
</script>
<script>
(function (i, s, o, g, r, a, m) {
i['GoogleAnalyticsObject'] = r; i[r] = i[r] || function () {
(i[r].q = i[r].q || []).push(arguments)
}, i[r].l = 1 * new Date(); a = s.createElement(o),
m = s.getElementsByTagName(o)[0]; a.async = 1; a.src = g; m.parentNode.insertBefore(a, m)
})(window, document, 'script', '//www.google-analytics.com/analytics.js', 'ga');
ga('create', 'UA-45999426-3', 'auto');
ga('send', 'pageview');
</script>
</body>
</html>