Kafka configuration inspector
Project description
kafkacfg
: Kafka configuration analyzer CLI
Kafka has a lot of parameters, tunables and knobs, and the configuration page isn't the easiest to parse. kafkacfg
allows you to analyze your kafka configuration file, to help you understand what it does.
[!NOTE] all
kafkacfg
commands (exceptkafkacfg recommends
) can be passed a--source librdkafka
argument, to work on thelibrdkafka
configuration parameters instead of the Kafka parameters
kafkacfg explain
This command is used to display extra information about each configuration tunable used in your configuration file.
$ cat samples/server.properties
num.io.threads = 10
broker.id = 1001
auto.create.topics.enable = false
background.threads = 10
$ kafkacfg explain -k 3.4 samples/server.properties | jq .
[
{
"name": "num.io.threads",
"override": "10",
"description": "The number of threads that the server uses for processing requests, which may include disk I/O",
"scope": "broker",
"type": "int",
"default": "8",
"valid_values": "[1,...]",
"importance": "high",
"update_mode": "cluster-wide"
},
{
"name": "broker.id",
"override": "1001",
"description": "The broker id for this server. If unset, a unique broker id will be generated.To avoid conflicts between zookeeper generated broker id's and user configured broker id's, generated broker ids start from reserved.broker.max.id + 1.",
"scope": "broker",
"type": "int",
"default": "-1",
"valid_values": null,
"importance": "high",
"update_mode": "read-only"
},
{
"name": "auto.create.topics.enable",
"override": "false",
"description": "Enable auto creation of topic on the server",
"scope": "broker",
"type": "boolean",
"default": "true",
"valid_values": null,
"importance": "high",
"update_mode": "read-only"
},
{
"name": "background.threads",
"override": "10",
"description": "The number of threads to use for various background processing tasks",
"scope": "broker",
"type": "int",
"default": "10",
"valid_values": "[1,...]",
"importance": "high",
"update_mode": "cluster-wide"
}
]
This format makes it easy to pipe it through table formatting tools such as jtbl
:
$ kafkacfg explain -k 3.4 samples/server.properties | jtbl
╒═══════════════════════════╤════════════╤═══════════════════════════════════════════════════════════════════════╤═════════╤═════════╤═══════════╤════════════════╤══════════════╤═══════════════╕
│ name │ override │ description │ scope │ type │ default │ valid_values │ importance │ update_mode │
╞═══════════════════════════╪════════════╪═══════════════════════════════════════════════════════════════════════╪═════════╪═════════╪═══════════╪════════════════╪══════════════╪═══════════════╡
│ num.io.threads │ 10 │ The number of threads that the server uses for processing requests, w │ broker │ int │ 8 │ [1,...] │ high │ cluster-wide │
│ │ │ hich may include disk I/O │ │ │ │ │ │ │
├───────────────────────────┼────────────┼───────────────────────────────────────────────────────────────────────┼─────────┼─────────┼───────────┼────────────────┼──────────────┼───────────────┤
│ broker.id │ 1001 │ The broker id for this server. If unset, a unique broker id will be g │ broker │ int │ -1 │ │ high │ read-only │
│ │ │ enerated.To avoid conflicts between zookeeper generated broker id's a │ │ │ │ │ │ │
│ │ │ nd user configured broker id's, generated broker ids start from reser │ │ │ │ │ │ │
│ │ │ ved.broker.max.id + 1. │ │ │ │ │ │ │
├───────────────────────────┼────────────┼───────────────────────────────────────────────────────────────────────┼─────────┼─────────┼───────────┼────────────────┼──────────────┼───────────────┤
│ auto.create.topics.enable │ false │ Enable auto creation of topic on the server │ broker │ boolean │ true │ │ high │ read-only │
├───────────────────────────┼────────────┼───────────────────────────────────────────────────────────────────────┼─────────┼─────────┼───────────┼────────────────┼──────────────┼───────────────┤
│ background.threads │ 10 │ The number of threads to use for various background processing tasks │ broker │ int │ 10 │ [1,...] │ high │ cluster-wide │
╘═══════════════════════════╧════════════╧═══════════════════════════════════════════════════════════════════════╧═════════╧═════════╧═══════════╧════════════════╧══════════════╧═══════════════╛
You can also explain producer and consumer level configurations, by using the -c / --config-type
flag:
$ cat -p producer.properties
compression.type=lz4
retries=2
$ kafkacfg explain -k 1.1 --config-type producer producer.properties | jtbl
╒══════════════════╤════════════╤═══════════════════════════════════════════════════════════════════════════════════════════════╤══════════╤════════╤═══════════╤════════════════════╤══════════════╕
│ name │ override │ description │ scope │ type │ default │ valid_values │ importance │
╞══════════════════╪════════════╪═══════════════════════════════════════════════════════════════════════════════════════════════╪══════════╪════════╪═══════════╪════════════════════╪══════════════╡
│ compression.type │ lz4 │ The compression type for all data generated by the producer. The default is none (i.e. no com │ producer │ string │ none │ │ high │
│ │ │ pression). Valid values are none, gzip, snappy, or lz4. Compression is of full batches of da │ │ │ │ │ │
│ │ │ ta, so the efficacy of batching will also impact the compression ratio (more batching means b │ │ │ │ │ │
│ │ │ etter compression). │ │ │ │ │ │
├──────────────────┼────────────┼───────────────────────────────────────────────────────────────────────────────────────────────┼──────────┼────────┼───────────┼────────────────────┼──────────────┤
│ retries │ 2 │ Setting a value greater than zero will cause the client to resend any record whose send fails │ producer │ int │ 0 │ [0,...,2147483647] │ high │
│ │ │ with a potentially transient error. Note that this retry is no different than if the client │ │ │ │ │ │
│ │ │ resent the record upon receiving the error. Allowing retries without setting max.in.flight.re │ │ │ │ │ │
│ │ │ quests.per.connection to 1 will potentially change the ordering of records because if two bat │ │ │ │ │ │
│ │ │ ches are sent to a single partition, and the first fails and is retried but the second succee │ │ │ │ │ │
│ │ │ ds, then the records in the second batch may appear first. │ │ │ │ │ │
╘══════════════════╧════════════╧═══════════════════════════════════════════════════════════════════════════════════════════════╧══════════╧════════╧═══════════╧════════════════════╧══════════════╛
kafkacfg overrides
The overrides
command will only display and enrich the configuration tunables with non-default values, to help you understand in what way your kafka configuration was tuned:
$ kafkacfg overrides -k 3.4 samples/server.properties | jtbl
name override description scope type default valid_values importance update_mode
------------------------- ---------- ---------------------------------------------------------------------------------------------- ------- ------- --------- -------------- ------------ -------------
num.io.threads 10 The number of threads that the server uses for processing requests, which may include disk I/O broker int 8 [1,...] high cluster-wide
auto.create.topics.enable false Enable auto creation of topic on the server broker boolean true high read-only
You can also explain producer and consumer level configurations override, by using the -c / --config-type
flag.
kafkacfg query
The query
command allows you to filter the kafka configuration tunables with a custom query, while enriching the filter results with the usual metadata and potential override value.
$ kafkacfg query -v 1.1 'name=replica.fetch.min.bytes' | jtbl
name override description scope type default valid_values importance update_mode
----------------------- ---------- ---------------------------------------------------------------------------------------------------- ------- ------ --------- -------------- ------------ -------------
replica.fetch.min.bytes Minimum bytes expected for each fetch response. If not enough bytes, wait up to replicaMaxWaitTimeMs broker int 1 high read-only
You can also compose filter predicates and even use *
for a glob match:
$ kafkacfg query -v 3.4 "name=replica.*.bytes;importance=high" | jtbl
╒═════════════════════════════════════╤════════════╤═══════════════════════════════════════════════════╤═════════╤════════╤══════════════════════╤════════════════╤══════════════╤═══════════════╕
│ name │ override │ description │ scope │ type │ default │ valid_values │ importance │ update_mode │
╞═════════════════════════════════════╪════════════╪═══════════════════════════════════════════════════╪═════════╪════════╪══════════════════════╪════════════════╪══════════════╪═══════════════╡
│ replica.fetch.min.bytes │ │ Minimum bytes expected for each fetch response. I │ broker │ int │ 1 │ │ high │ read-only │
│ │ │ f not enough bytes, wait up to replica.fetch.wait │ │ │ │ │ │ │
│ │ │ .max.ms (broker config). │ │ │ │ │ │ │
├─────────────────────────────────────┼────────────┼───────────────────────────────────────────────────┼─────────┼────────┼──────────────────────┼────────────────┼──────────────┼───────────────┤
│ replica.socket.receive.buffer.bytes │ │ The socket receive buffer for network requests │ broker │ int │ 65536 (64 kibibytes) │ │ high │ read-only │
╘═════════════════════════════════════╧════════════╧═══════════════════════════════════════════════════╧═════════╧════════╧══════════════════════╧════════════════╧══════════════╧═══════════════╛
kafkacfg recommends
The recommends
command emits sourced configuration recommendation based on broker attributes (number of CPUs, number of disks, etc).
Note: These are suggestions more than absolute thruths. We encourage you to read the sources, test the configuration values and make your own mind.
$ kafkacfg recommends -k 3.4 --broker-num-cpus 24 --broker-num-disks 12 server.properties
Recommendation 1: set num.io.threads == 12 (num_disks)
* https://strimzi.io/blog/2021/06/08/broker-tuning/
Recommendation 2: set replica.fetch.min.bytes == 512
* https://azure.microsoft.com/en-us/blog/processing-trillions-of-events-per-day-with-apache-kafka-on-azure/
Recommendation 3: set socket.receive.buffer.bytes == 512
* https://azure.microsoft.com/en-us/blog/processing-trillions-of-events-per-day-with-apache-kafka-on-azure/
* https://strimzi.io/blog/2021/06/08/broker-tuning/
Recommendation 4: set socket.send.buffer.bytes == 1048576
* https://azure.microsoft.com/en-us/blog/processing-trillions-of-events-per-day-with-apache-kafka-on-azure/
* https://strimzi.io/blog/2021/06/08/broker-tuning/
Recommendation 5: set replica.socket.receive.buffer.bytes == 1048576
* https://azure.microsoft.com/en-us/blog/processing-trillions-of-events-per-day-with-apache-kafka-on-azure/
kafkacfg diff
The diff
command displays what configuration tunables have changed between 2 kafka versions, whether they have been deprecated, newly introduced, or their default value has changed.
$ kafkacfg diff --from-version 3.5 --to-version 3.6 | jtbl
╒═══════════════════════════════════════════════╤══════════╤════════════════════════════════════════════════════════════╤═════════╤═════════╤════════════════════════════════════════════════════════════╤══════════════╤═══════════════╤═════════════════╤════════════════════════════════════════════════════════════╤═════════════════════════════════════╕
│ name │ status │ description │ scope │ type │ valid_values │ importance │ update_mode │ default (3.5) │ default (3.6) │ server_default_property │
╞═══════════════════════════════════════════════╪══════════╪════════════════════════════════════════════════════════════╪═════════╪═════════╪════════════════════════════════════════════════════════════╪══════════════╪═══════════════╪═════════════════╪════════════════════════════════════════════════════════════╪═════════════════════════════════════╡
│ log.local.retention.bytes │ new │ The maximum size of local log segments that can grow for a │ broker │ long │ [-2,...] │ medium │ cluster-wide │ [ABSENT] │ -2 │ │
│ │ │ partition before it gets eligible for deletion. Default v │ │ │ │ │ │ │ │ │
│ │ │ alue is -2, it represents `log.retention.bytes` value to b │ │ │ │ │ │ │ │ │
│ │ │ e used. The effective value should always be less than or │ │ │ │ │ │ │ │ │
│ │ │ equal to `log.retention.bytes` value. │ │ │ │ │ │ │ │ │
├───────────────────────────────────────────────┼──────────┼────────────────────────────────────────────────────────────┼─────────┼─────────┼────────────────────────────────────────────────────────────┼──────────────┼───────────────┼─────────────────┼────────────────────────────────────────────────────────────┼─────────────────────────────────────┤
│ log.local.retention.ms │ new │ The number of milliseconds to keep the local log segments │ broker │ long │ [-2,...] │ medium │ cluster-wide │ [ABSENT] │ -2 │ │
│ │ │ before it gets eligible for deletion. Default value is -2, │ │ │ │ │ │ │ │ │
│ │ │ it represents `log.retention.ms` value is to be used. The │ │ │ │ │ │ │ │ │
│ │ │ effective value should always be less than or equal to `l │ │ │ │ │ │ │ │ │
│ │ │ og.retention.ms` value. │ │ │ │ │ │ │ │ │
├───────────────────────────────────────────────┼──────────┼────────────────────────────────────────────────────────────┼─────────┼─────────┼────────────────────────────────────────────────────────────┼──────────────┼───────────────┼─────────────────┼────────────────────────────────────────────────────────────┼─────────────────────────────────────┤
│ log.message.timestamp.after.max.ms │ new │ This configuration sets the allowable timestamp difference │ broker │ long │ [0,...] │ medium │ cluster-wide │ [ABSENT] │ 9223372036854775807 │ │
│ │ │ between the message timestamp and the broker's timestamp. │ │ │ │ │ │ │ │ │
│ │ │ The message timestamp can be later than or equal to the b │ │ │ │ │ │ │ │ │
│ │ │ roker's timestamp, with the maximum allowable difference d │ │ │ │ │ │ │ │ │
│ │ │ etermined by the value set in this configuration. If log.m │ │ │ │ │ │ │ │ │
│ │ │ essage.timestamp.type=CreateTime, the message will be reje │ │ │ │ │ │ │ │ │
│ │ │ cted if the difference in timestamps exceeds this specifie │ │ │ │ │ │ │ │ │
│ │ │ d threshold. This configuration is ignored if log.message. │ │ │ │ │ │ │ │ │
│ │ │ timestamp.type=LogAppendTime. │ │ │ │ │ │ │ │ │
├───────────────────────────────────────────────┼──────────┼────────────────────────────────────────────────────────────┼─────────┼─────────┼────────────────────────────────────────────────────────────┼──────────────┼───────────────┼─────────────────┼────────────────────────────────────────────────────────────┼─────────────────────────────────────┤
...
You can also highlight changes in producer and consumer level configurations, by using the -c / --config-type
flag.
Project details
Download files
Download the file for your platform. If you're not sure which to choose, learn more about installing packages.