risingwave_connector/source/kafka/
mod.rs

1// Copyright 2025 RisingWave Labs
2//
3// Licensed under the Apache License, Version 2.0 (the "License");
4// you may not use this file except in compliance with the License.
5// You may obtain a copy of the License at
6//
7//     http://www.apache.org/licenses/LICENSE-2.0
8//
9// Unless required by applicable law or agreed to in writing, software
10// distributed under the License is distributed on an "AS IS" BASIS,
11// WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
12// See the License for the specific language governing permissions and
13// limitations under the License.
14
15use std::collections::HashMap;
16
17use serde::Deserialize;
18use serde_with::{DisplayFromStr, serde_as};
19
20use crate::connector_common::{AwsAuthProps, KafkaConnectionProps, KafkaPrivateLinkCommon};
21use crate::enforce_secret::EnforceSecret;
22use crate::error::ConnectorResult;
23
24mod client_context;
25pub mod enumerator;
26pub mod private_link;
27pub mod source;
28pub mod split;
29pub mod stats;
30
31pub use client_context::*;
32pub use enumerator::*;
33pub use source::*;
34pub use split::*;
35use with_options::WithOptions;
36
37use crate::connector_common::{KafkaCommon, RdKafkaPropertiesCommon};
38use crate::source::SourceProperties;
39
40pub const KAFKA_CONNECTOR: &str = "kafka";
41pub const KAFKA_PROPS_BROKER_KEY: &str = "properties.bootstrap.server";
42pub const KAFKA_PROPS_BROKER_KEY_ALIAS: &str = "kafka.brokers";
43pub const PRIVATELINK_CONNECTION: &str = "privatelink";
44
45/// Properties for the rdkafka library. Leave a field as `None` to use the default value.
46/// These properties are not intended to be exposed to users in the majority of cases.
47///
48/// See also <https://github.com/confluentinc/librdkafka/blob/master/CONFIGURATION.md>
49#[serde_as]
50#[derive(Clone, Debug, Deserialize, Default, WithOptions)]
51pub struct RdKafkaPropertiesConsumer {
52    /// Minimum number of messages per topic+partition librdkafka tries to maintain in the local
53    /// consumer queue.
54    #[serde(rename = "properties.queued.min.messages")]
55    #[serde_as(as = "Option<DisplayFromStr>")]
56    pub queued_min_messages: Option<usize>,
57
58    #[serde(rename = "properties.queued.max.messages.kbytes")]
59    #[serde_as(as = "Option<DisplayFromStr>")]
60    pub queued_max_messages_kbytes: Option<usize>,
61
62    /// Maximum time the broker may wait to fill the Fetch response with `fetch.min.`bytes of
63    /// messages.
64    #[serde(rename = "properties.fetch.wait.max.ms")]
65    #[serde_as(as = "Option<DisplayFromStr>")]
66    pub fetch_wait_max_ms: Option<usize>,
67
68    /// Minimum number of bytes the broker responds with. If fetch.wait.max.ms expires the accumulated data will be sent to the client regardless of this setting.
69    #[serde(rename = "properties.fetch.min.bytes")]
70    #[serde_as(as = "Option<DisplayFromStr>")]
71    pub fetch_min_bytes: Option<usize>,
72
73    /// Initial maximum number of bytes per topic+partition to request when fetching messages from the broker. If the client encounters a message larger than this value it will gradually try to increase it until the entire message can be fetched.
74    #[serde(rename = "properties.fetch.message.max.bytes")]
75    #[serde_as(as = "Option<DisplayFromStr>")]
76    pub fetch_message_max_bytes: Option<usize>,
77
78    /// How long to postpone the next fetch request for a topic+partition in case the current fetch
79    /// queue thresholds (`queued.min.messages` or `queued.max.messages.kbytes`) have been
80    /// exceeded. This property may need to be decreased if the queue thresholds are set low
81    /// and the application is experiencing long (~1s) delays between messages. Low values may
82    /// increase CPU utilization.
83    #[serde(rename = "properties.fetch.queue.backoff.ms")]
84    #[serde_as(as = "Option<DisplayFromStr>")]
85    pub fetch_queue_backoff_ms: Option<usize>,
86
87    /// Maximum amount of data the broker shall return for a Fetch request. Messages are fetched in
88    /// batches by the consumer and if the first message batch in the first non-empty partition of
89    /// the Fetch request is larger than this value, then the message batch will still be returned
90    /// to ensure the consumer can make progress. The maximum message batch size accepted by the
91    /// broker is defined via `message.max.bytes` (broker config) or `max.message.bytes` (broker
92    /// topic config). `fetch.max.bytes` is automatically adjusted upwards to be at least
93    /// `message.max.bytes` (consumer config).
94    #[serde(rename = "properties.fetch.max.bytes")]
95    #[serde_as(as = "Option<DisplayFromStr>")]
96    pub fetch_max_bytes: Option<usize>,
97
98    /// Whether to automatically and periodically commit offsets in the background.
99    ///
100    /// Note that RisingWave does NOT rely on committed offsets. Committing offset is only for exposing the
101    /// progress for monitoring. Setting this to false can avoid creating consumer groups.
102    ///
103    /// default: true
104    #[serde(rename = "properties.enable.auto.commit")]
105    #[serde_as(as = "Option<DisplayFromStr>")]
106    pub enable_auto_commit: Option<bool>,
107}
108
109#[derive(Clone, Debug, Deserialize, WithOptions)]
110pub struct KafkaProperties {
111    /// This parameter is not intended to be exposed to users.
112    /// This parameter specifies only for one parallelism. The parallelism of kafka source
113    /// is equal to the parallelism passed into compute nodes. So users need to calculate
114    /// how many bytes will be consumed in total across all the parallelism by themselves.
115    #[serde(rename = "bytes.per.second", alias = "kafka.bytes.per.second")]
116    pub bytes_per_second: Option<String>,
117
118    /// This parameter is not intended to be exposed to users.
119    /// This parameter specifies only for one parallelism. The parallelism of kafka source
120    /// is equal to the parallelism passed into compute nodes. So users need to calculate
121    /// how many messages will be consumed in total across all the parallelism by themselves.
122    #[serde(rename = "max.num.messages", alias = "kafka.max.num.messages")]
123    pub max_num_messages: Option<String>,
124
125    #[serde(rename = "scan.startup.mode", alias = "kafka.scan.startup.mode")]
126    pub scan_startup_mode: Option<String>,
127
128    #[serde(
129        rename = "scan.startup.timestamp.millis",
130        alias = "kafka.time.offset",
131        alias = "scan.startup.timestamp_millis" // keep for compatibility
132    )]
133    pub time_offset: Option<String>,
134
135    /// Specify a custom consumer group id prefix for the source.
136    /// Defaults to `rw-consumer`.
137    ///
138    /// Notes:
139    /// - Each job (materialized view) will have a separated consumer group and
140    ///   contains a generated suffix in the group id.
141    ///   The consumer group will be `{group_id_prefix}-{fragment_id}`.
142    /// - The consumer group is solely for monintoring progress in some external
143    ///   Kafka tools, and for authorization. RisingWave does not rely on committed
144    ///   offsets, and does not join the consumer group. It just reports offsets
145    ///   to the group.
146    #[serde(rename = "group.id.prefix")]
147    pub group_id_prefix: Option<String>,
148
149    /// This parameter is used to tell `KafkaSplitReader` to produce `UpsertMessage`s, which
150    /// combine both key and value fields of the Kafka message.
151    /// TODO: Currently, `Option<bool>` can not be parsed here.
152    #[serde(rename = "upsert")]
153    pub upsert: Option<String>,
154
155    #[serde(flatten)]
156    pub common: KafkaCommon,
157
158    #[serde(flatten)]
159    pub connection: KafkaConnectionProps,
160
161    #[serde(flatten)]
162    pub rdkafka_properties_common: RdKafkaPropertiesCommon,
163
164    #[serde(flatten)]
165    pub rdkafka_properties_consumer: RdKafkaPropertiesConsumer,
166
167    #[serde(flatten)]
168    pub privatelink_common: KafkaPrivateLinkCommon,
169
170    #[serde(flatten)]
171    pub aws_auth_props: AwsAuthProps,
172
173    #[serde(flatten)]
174    pub unknown_fields: HashMap<String, String>,
175}
176
177impl EnforceSecret for KafkaProperties {
178    fn enforce_secret<'a>(prop_iter: impl Iterator<Item = &'a str>) -> ConnectorResult<()> {
179        for prop in prop_iter {
180            KafkaConnectionProps::enforce_one(prop)?;
181            AwsAuthProps::enforce_one(prop)?;
182        }
183        Ok(())
184    }
185}
186
187impl SourceProperties for KafkaProperties {
188    type Split = KafkaSplit;
189    type SplitEnumerator = KafkaSplitEnumerator;
190    type SplitReader = KafkaSplitReader;
191
192    const SOURCE_NAME: &'static str = KAFKA_CONNECTOR;
193}
194
195impl crate::source::UnknownFields for KafkaProperties {
196    fn unknown_fields(&self) -> HashMap<String, String> {
197        self.unknown_fields.clone()
198    }
199}
200
201impl KafkaProperties {
202    pub fn set_client(&self, c: &mut rdkafka::ClientConfig) {
203        self.rdkafka_properties_common.set_client(c);
204        self.rdkafka_properties_consumer.set_client(c);
205    }
206
207    pub fn group_id(&self, fragment_id: u32) -> String {
208        format!(
209            "{}-{}",
210            self.group_id_prefix.as_deref().unwrap_or("rw-consumer"),
211            fragment_id
212        )
213    }
214}
215
216const KAFKA_ISOLATION_LEVEL: &str = "read_committed";
217
218impl RdKafkaPropertiesConsumer {
219    pub fn set_client(&self, c: &mut rdkafka::ClientConfig) {
220        if let Some(v) = &self.queued_min_messages {
221            c.set("queued.min.messages", v.to_string());
222        }
223        if let Some(v) = &self.queued_max_messages_kbytes {
224            c.set("queued.max.messages.kbytes", v.to_string());
225        }
226        if let Some(v) = &self.fetch_wait_max_ms {
227            c.set("fetch.wait.max.ms", v.to_string());
228        }
229        if let Some(v) = &self.fetch_min_bytes {
230            c.set("fetch.min.bytes", v.to_string());
231        }
232        if let Some(v) = &self.fetch_message_max_bytes {
233            c.set("fetch.message.max.bytes", v.to_string());
234        }
235        if let Some(v) = &self.fetch_queue_backoff_ms {
236            c.set("fetch.queue.backoff.ms", v.to_string());
237        }
238        if let Some(v) = &self.fetch_max_bytes {
239            c.set("fetch.max.bytes", v.to_string());
240        }
241        if let Some(v) = &self.enable_auto_commit {
242            c.set("enable.auto.commit", v.to_string());
243        }
244    }
245}
246
247#[cfg(test)]
248mod test {
249    use std::collections::BTreeMap;
250
251    use maplit::btreemap;
252
253    use super::*;
254
255    #[test]
256    fn test_parse_config_consumer_common() {
257        let config: BTreeMap<String, String> = btreemap! {
258            // common
259            "properties.bootstrap.server".to_owned() => "127.0.0.1:9092".to_owned(),
260            "topic".to_owned() => "test".to_owned(),
261            // kafka props
262            "scan.startup.mode".to_owned() => "earliest".to_owned(),
263            // RdKafkaPropertiesCommon
264            "properties.message.max.bytes".to_owned() => "12345".to_owned(),
265            "properties.receive.message.max.bytes".to_owned() => "54321".to_owned(),
266            // RdKafkaPropertiesConsumer
267            "properties.queued.min.messages".to_owned() => "114514".to_owned(),
268            "properties.queued.max.messages.kbytes".to_owned() => "114514".to_owned(),
269            "properties.fetch.wait.max.ms".to_owned() => "114514".to_owned(),
270            "properties.fetch.max.bytes".to_owned() => "114514".to_owned(),
271            "properties.enable.auto.commit".to_owned() => "true".to_owned(),
272            "properties.fetch.queue.backoff.ms".to_owned() => "114514".to_owned(),
273            // PrivateLink
274            "broker.rewrite.endpoints".to_owned() => "{\"broker1\": \"10.0.0.1:8001\"}".to_owned(),
275        };
276
277        let props: KafkaProperties =
278            serde_json::from_value(serde_json::to_value(config).unwrap()).unwrap();
279
280        assert_eq!(props.scan_startup_mode, Some("earliest".to_owned()));
281        assert_eq!(
282            props.rdkafka_properties_common.receive_message_max_bytes,
283            Some(54321)
284        );
285        assert_eq!(
286            props.rdkafka_properties_common.message_max_bytes,
287            Some(12345)
288        );
289        assert_eq!(
290            props.rdkafka_properties_consumer.queued_min_messages,
291            Some(114514)
292        );
293        assert_eq!(
294            props.rdkafka_properties_consumer.queued_max_messages_kbytes,
295            Some(114514)
296        );
297        assert_eq!(
298            props.rdkafka_properties_consumer.fetch_wait_max_ms,
299            Some(114514)
300        );
301        assert_eq!(
302            props.rdkafka_properties_consumer.fetch_max_bytes,
303            Some(114514)
304        );
305        assert_eq!(
306            props.rdkafka_properties_consumer.enable_auto_commit,
307            Some(true)
308        );
309        assert_eq!(
310            props.rdkafka_properties_consumer.fetch_queue_backoff_ms,
311            Some(114514)
312        );
313        let hashmap: BTreeMap<String, String> = btreemap! {
314            "broker1".to_owned() => "10.0.0.1:8001".to_owned()
315        };
316        assert_eq!(props.privatelink_common.broker_rewrite_map, Some(hashmap));
317    }
318}
risingwave_connector/source/kafka/mod.rs

risingwave_connector/source/kafka/
mod.rs
