/
StructuredComplexSessionization.scala
301 lines (262 loc) · 10.8 KB
/
StructuredComplexSessionization.scala
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
/*
* Licensed to the Apache Software Foundation (ASF) under one or more
* contributor license agreements. See the NOTICE file distributed with
* this work for additional information regarding copyright ownership.
* The ASF licenses this file to You under the Apache License, Version 2.0
* (the "License"); you may not use this file except in compliance with
* the License. You may obtain a copy of the License at
*
* http://www.apache.org/licenses/LICENSE-2.0
*
* Unless required by applicable law or agreed to in writing, software
* distributed under the License is distributed on an "AS IS" BASIS,
* WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
* See the License for the specific language governing permissions and
* limitations under the License.
*/
// scalastyle:off println
package org.apache.spark.examples.sql.streaming
import java.sql.Timestamp
import scala.collection.mutable
import org.apache.spark.sql.SparkSession
import org.apache.spark.sql.functions._
import org.apache.spark.sql.streaming._
import org.apache.spark.sql.types.{StringType, StructType, TimestampType}
/**
* Sessionize events in UTF8 encoded, '\n' delimited text received from the network.
* Each line composes an event, and the line should match to the json format.
*
* The schema of the event is following:
* - user_id: String
* - event_type: String
* - timestamp: Long
*
* The supported types are following:
* - NEW_EVENT
* - CLOSE_SESSION
*
* This example focuses to demonstrate the complex sessionization which uses two conditions
* on closing session; conditions are following:
* - No further event is provided for the user ID within 5 seconds
* - An event having CLOSE_SESSION as event_type is provided for the user ID
*
* Usage: StructuredComplexSessionization <hostname> <port>
* <hostname> and <port> describe the TCP server that Structured Streaming
* would connect to receive data.
*
* To run this on your local machine, you need to first run a Netcat server
* `$ nc -lk 9999`
* and then run the example
* `$ bin/run-example sql.streaming.StructuredComplexSessionization
* localhost 9999`
*
* Here's a set of events for example:
*
* {"user_id": "user1", "event_type": "NEW_EVENT", "timestamp": 13}
* {"user_id": "user1", "event_type": "NEW_EVENT", "timestamp": 10}
* {"user_id": "user1", "event_type": "CLOSE_SESSION", "timestamp": 15}
* {"user_id": "user1", "event_type": "NEW_EVENT", "timestamp": 17}
* {"user_id": "user1", "event_type": "NEW_EVENT", "timestamp": 19}
* {"user_id": "user1", "event_type": "NEW_EVENT", "timestamp": 29}
*
* {"user_id": "user2", "event_type": "NEW_EVENT", "timestamp": 45}
*
* {"user_id": "user1", "event_type": "NEW_EVENT", "timestamp": 65}
*
* and results (the output can be split across micro-batches):
*
* +-----+----------+---------+
* | id|durationMs|numEvents|
* +-----+----------+---------+
* |user1| 5000| 3|
* |user1| 7000| 2|
* |user1| 5000| 1|
* |user2| 5000| 1|
* +-----+----------+---------+
* (The last event is not reflected into output due to watermark.)
*
* Note that there're three different sessions for 'user1'. The events in first two sessions
* are occurred within gap duration for nearest events, but they don't compose a single session
* due to the event of CLOSE_SESSION.
*
* Also note that the implementation is simplified one. This example doesn't address
* - UPDATE MODE (the semantic is not clear for session window with event time processing)
* - partial merge (events in session which are earlier than watermark can be aggregated)
* - other possible optimizations
*/
object StructuredComplexSessionization {
def main(args: Array[String]): Unit = {
if (args.length < 2) {
System.err.println("Usage: StructuredComplexSessionization <hostname> <port>")
System.exit(1)
}
val host = args(0)
val port = args(1).toInt
val spark = SparkSession
.builder
.appName("StructuredComplexSessionization")
.getOrCreate()
import spark.implicits._
// Create DataFrame representing the stream of input lines from connection to host:port
val lines = spark.readStream
.format("socket")
.option("host", host)
.option("port", port)
.load()
val jsonSchema = new StructType()
.add("user_id", StringType, nullable = false)
.add("event_type", StringType, nullable = false)
.add("timestamp", TimestampType, nullable = false)
val gapDuration: Long = 5 * 1000 // 5 seconds
// Parse the line into event, as described in classdoc.
val events = lines
.select(from_json(col("value"), jsonSchema).as("event"))
.selectExpr("event.user_id AS user_id", "event.event_type AS event_type",
"event.timestamp AS timestamp")
.withWatermark("timestamp", "10 seconds")
.as[(String, String, Timestamp)]
// Sessionize the events. Track number of events, start and end timestamps of session,
// and report session when session is closed.
val sessionUpdates = events
.groupByKey(event => event._1)
.flatMapGroupsWithState[List[SessionAcc], Session](OutputMode.Append(),
GroupStateTimeout.EventTimeTimeout) {
case (userId: String, events: Iterator[(String, String, Timestamp)],
state: GroupState[List[SessionAcc]]) =>
def handleEvict(sessions: List[SessionAcc]): Iterator[Session] = {
// we sorted sessions by timestamp
val (evicted, kept) = sessions.span {
s => s.endTime.getTime < state.getCurrentWatermarkMs()
}
if (kept.isEmpty) {
state.remove()
} else {
state.update(kept)
// trigger timeout at the end time of the first session
state.setTimeoutTimestamp(kept.head.endTime.getTime)
}
evicted.map { sessionAcc =>
Session(userId, sessionAcc.endTime.getTime - sessionAcc.startTime.getTime,
sessionAcc.events.length)
}.iterator
}
def mergeSessions(sessions: List[SessionAcc]): Unit = {
// we sorted sessions by timestamp
val updatedSessions = new mutable.ArrayBuffer[SessionAcc]()
updatedSessions ++= sessions
var curIdx = 0
while (curIdx < updatedSessions.length - 1) {
val curSession = updatedSessions(curIdx)
val nextSession = updatedSessions(curIdx + 1)
// Current session and next session can be merged
if (curSession.endTime.getTime > nextSession.startTime.getTime) {
val accumulatedEvents =
(curSession.events ++ nextSession.events).sortBy(_.startTimestamp.getTime)
val newSessions = new mutable.ArrayBuffer[SessionAcc]()
var eventsForCurSession = new mutable.ArrayBuffer[SessionEvent]()
accumulatedEvents.foreach { event =>
eventsForCurSession += event
if (event.eventType == EventTypes.CLOSE_SESSION) {
newSessions += SessionAcc(eventsForCurSession.toList)
eventsForCurSession = new mutable.ArrayBuffer[SessionEvent]()
}
}
if (eventsForCurSession.nonEmpty) {
newSessions += SessionAcc(eventsForCurSession.toList)
}
// replace current session and next session with new session(s)
updatedSessions.remove(curIdx + 1)
updatedSessions(curIdx) = newSessions.head
if (newSessions.length > 1) {
updatedSessions.insertAll(curIdx + 1, newSessions.tail)
}
// move the cursor to the last new session(s)
curIdx += newSessions.length - 1
} else {
// move to the next session
curIdx += 1
}
}
// update state
state.update(updatedSessions.toList)
}
if (state.hasTimedOut && state.exists) {
handleEvict(state.get.sortBy(_.startTime.getTime))
} else {
// convert each event as individual session
val sessionsFromEvents = events.map { case (userId, eventType, timestamp) =>
val e = SessionEvent(userId, eventType, timestamp, gapDuration)
SessionAcc(List(e))
}.toList
if (sessionsFromEvents.nonEmpty) {
val sessionsFromState = if (state.exists) {
state.get
} else {
List.empty
}
// sort sessions via start timestamp, and merge
mergeSessions((sessionsFromEvents ++ sessionsFromState).sortBy(_.startTime.getTime))
// we still need to handle eviction here
handleEvict(state.get.sortBy(_.startTime.getTime))
} else {
Iterator.empty
}
}
}
// Start running the query that prints the session updates to the console
val query = sessionUpdates
.writeStream
.outputMode("append")
.format("console")
.start()
query.awaitTermination()
}
}
object EventTypes extends Enumeration {
type EventTypes = Value
val NEW_EVENT, CLOSE_SESSION = Value
}
case class SessionEvent(
userId: String,
eventType: EventTypes.Value,
startTimestamp: Timestamp,
endTimestamp: Timestamp)
object SessionEvent {
def apply(
userId: String,
eventTypeStr: String,
timestamp: Timestamp,
gapDuration: Long): SessionEvent = {
val eventType = EventTypes.withName(eventTypeStr)
val endTime = if (eventType == EventTypes.CLOSE_SESSION) {
timestamp
} else {
new Timestamp(timestamp.getTime + gapDuration)
}
SessionEvent(userId, eventType, timestamp, endTime)
}
}
case class SessionAcc(events: List[SessionEvent]) {
private val sortedEvents: List[SessionEvent] = events.sortBy(_.startTimestamp.getTime)
require(!sortedEvents.dropRight(1).exists(_.eventType == EventTypes.CLOSE_SESSION),
"CLOSE_SESSION event cannot be placed except the last event!")
def eventsAsSorted: List[SessionEvent] = sortedEvents
def startTime: Timestamp = sortedEvents.head.startTimestamp
def endTime: Timestamp = sortedEvents.last.endTimestamp
override def toString: String = {
s"SessionAcc(events: $events / sorted: $sortedEvents / " +
s"start time: $startTime / endTime: $endTime)"
}
}
/**
* User-defined data type representing the session information returned by flatMapGroupsWithState.
*
* @param id Id of the user
* @param durationMs Duration the session was active, that is, from first event to its expiry
* @param numEvents Number of events received by the session while it was active
*/
case class Session(
id: String,
durationMs: Long,
numEvents: Int)
// scalastyle:on println