File: eos-transactions.py

package info (click to toggle)
python-confluent-kafka 1.7.0-4
  • links: PTS, VCS
  • area: main
  • in suites: bookworm, sid, trixie
  • size: 1,900 kB
  • sloc: python: 8,335; ansic: 6,065; sh: 1,203; makefile: 178
file content (200 lines) | stat: -rwxr-xr-x 6,962 bytes parent folder | download
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
#!/usr/bin/env python
# -*- coding: utf-8 -*-
#
# Copyright 2020 Confluent Inc.
#
# Licensed 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.


# This is a simple example of a `read-process-write` application
# using Apache Kafka's transactional API.
#
# See the following blog for additional information:
# https://www.confluent.io/blog/transactions-apache-kafka/
#


"""
The following example demonstrates how to perform a consume-transform-produce
loop with exactly-once semantics.

In order to achieve exactly-once semantics we use the transactional producer
and a single transaction aware consumer.

The following assumptions apply to the source data (input_topic below):
    1. There are no duplicates in the input topic.

##  A quick note about exactly-once-processing guarantees and Kafka. ##

The exactly once, and idempotence, guarantees start after the producer has been
provided a record. There is no way for a producer to identify a record as a
duplicate in isolation. Instead it is the application's job to ensure that only
a single copy of any record is passed to the producer.

Special care needs to be taken when expanding the consumer group to multiple
members.
Review KIP-447 for complete details.
"""

import argparse
from base64 import b64encode
from uuid import uuid4

from confluent_kafka import Producer, Consumer, KafkaError, TopicPartition


def process_input(msg):
    """
    Base64 encodes msg key/value contents
    :param msg:
    :returns: transformed key, value
    :rtype: tuple
    """

    key, value = None, None
    if msg.key() is not None:
        key = b64encode(msg.key())
    if msg.value() is not None:
        value = b64encode(msg.value())

    return key, value


def delivery_report(err, msg):
    """
    Reports message delivery status; success or failure
    :param KafkaError err: reason for delivery failure
    :param Message msg:
    :returns: None
    """
    if err:
        print('Message delivery failed ({} [{}]): {}'.format(
            msg.topic(), str(msg.partition()), err))


def main(args):
    brokers = args.brokers
    group_id = args.group_id
    input_topic = args.input_topic
    input_partition = args.input_partition
    output_topic = args.output_topic

    consumer = Consumer({
        'bootstrap.servers': brokers,
        'group.id': group_id,
        'auto.offset.reset': 'earliest',
        # Do not advance committed offsets outside of the transaction.
        # Consumer offsets are committed along with the transaction
        # using the producer's send_offsets_to_transaction() API.
        'enable.auto.commit': False,
        'enable.partition.eof': True,
    })

    # Prior to KIP-447 being supported each input partition requires
    # its own transactional producer, so in this example we use
    # assign() to a single partition rather than subscribe().
    # A more complex alternative is to dynamically create a producer per
    # partition in subscribe's rebalance callback.
    consumer.assign([TopicPartition(input_topic, input_partition)])

    producer = Producer({
        'bootstrap.servers': brokers,
        'transactional.id': 'eos-transactions.py'
    })

    # Initialize producer transaction.
    producer.init_transactions()
    # Start producer transaction.
    producer.begin_transaction()

    eof = {}
    msg_cnt = 0
    print("=== Starting Consume-Transform-Process loop ===")
    while True:
        # serve delivery reports from previous produce()s
        producer.poll(0)

        # read message from input_topic
        msg = consumer.poll(timeout=1.0)
        if msg is None:
            continue

        topic, partition = msg.topic(), msg.partition()
        if msg.error():
            if msg.error().code() == KafkaError._PARTITION_EOF:
                eof[(topic, partition)] = True
                print("=== Reached the end of {} [{}] at {}====".format(
                    topic, partition, msg.offset()))

                if len(eof) == len(consumer.assignment()):
                    print("=== Reached end of input ===")
                    break
            continue
        # clear EOF if a new message has been received
        eof.pop((topic, partition), None)

        msg_cnt += 1

        # process message
        processed_key, processed_value = process_input(msg)

        # produce transformed message to output topic
        producer.produce(output_topic, processed_value, processed_key,
                         on_delivery=delivery_report)

        if msg_cnt % 100 == 0:
            print("=== Committing transaction with {} messages at input offset {} ===".format(
                msg_cnt, msg.offset()))
            # Send the consumer's position to transaction to commit
            # them along with the transaction, committing both
            # input and outputs in the same transaction is what provides EOS.
            producer.send_offsets_to_transaction(
                consumer.position(consumer.assignment()),
                consumer.consumer_group_metadata())

            # Commit the transaction
            producer.commit_transaction()

            # Begin new transaction
            producer.begin_transaction()
            msg_cnt = 0

    print("=== Committing final transaction with {} messages ===".format(msg_cnt))
    # commit processed message offsets to the transaction
    producer.send_offsets_to_transaction(
        consumer.position(consumer.assignment()),
        consumer.consumer_group_metadata())

    # commit transaction
    producer.commit_transaction()

    consumer.close()


if __name__ == "__main__":
    parser = argparse.ArgumentParser(
        description="Exactly Once Semantics (EOS) example")
    parser.add_argument('-b', dest="brokers", required=True,
                        help="Bootstrap broker(s) (host[:port])")
    parser.add_argument('-t', dest="input_topic", required=True,
                        help="Input topic to consume from, an external " +
                        "producer needs to produce messages to this topic")
    parser.add_argument('-o', dest="output_topic", default="output_topic",
                        help="Output topic")
    parser.add_argument('-p', dest="input_partition", default=0, type=int,
                        help="Input partition to consume from")
    parser.add_argument('-g', dest="group_id",
                        default="eos_example_" + str(uuid4()),
                        help="Consumer group")

    main(parser.parse_args())