Kafka

The kafka output type writes a batch of messages to Kafka brokers and waits for acknowledgment before propagating it back to the input.

Common

# Common config fields, showing default values
output:
  label: ""
  kafka:
    addresses: [] # No default (required)
    topic: "" # No default (required)
    target_version: 2.1.0 # No default (optional)
    key: ""
    partitioner: fnv1a_hash
    compression: none
    static_headers: {} # No default (optional)
    metadata:
      exclude_prefixes: []
    max_in_flight: 64
    batching:
      count: 0
      byte_size: 0
      period: ""
      check: ""

Advanced

# All config fields, showing default values
output:
  label: ""
  kafka:
    addresses: [] # No default (required)
    tls:
      enabled: false
      skip_cert_verify: false
      enable_renegotiation: false
      root_cas: ""
      root_cas_file: ""
      client_certs: []
    sasl:
      mechanism: none
      user: ""
      password: ""
      access_token: ""
      token_cache: ""
      token_key: ""
    topic: "" # No default (required)
    client_id: tyk
    target_version: 2.1.0 # No default (optional)
    rack_id: ""
    key: ""
    partitioner: fnv1a_hash
    partition: ""
    custom_topic_creation:
      enabled: false
      partitions: -1
      replication_factor: -1
    compression: none
    static_headers: {} # No default (optional)
    metadata:
      exclude_prefixes: []
    inject_tracing_map: meta = @.merge(this) # No default (optional)
    max_in_flight: 64
    idempotent_write: false
    ack_replicas: false
    max_msg_bytes: 1000000
    timeout: 5s
    retry_as_batch: false
    batching:
      count: 0
      byte_size: 0
      period: ""
      check: ""
      processors: [] # No default (optional)
    max_retries: 0
    backoff:
      initial_interval: 3s
      max_interval: 10s
      max_elapsed_time: 30s

The config field ack_replicas determines whether we wait for acknowledgment from all replicas or just a single broker.

Metadata will be added to each message sent as headers (version 0.11+), but can be restricted using the field metadata.

Strict Ordering and Retries

When strict ordering is required for messages written to topic partitions it is important to ensure that both the field max_in_flight is set to 1 and that the field retry_as_batch is set to true.

You must also ensure that failed batches are never rerouted back to the same output. This can be done by setting the field max_retries to 0 and backoff.max_elapsed_time to empty, which will apply back pressure indefinitely until the batch is sent successfully.

However, this also means that manual intervention will eventually be required in cases where the batch cannot be sent due to configuration problems such as an incorrect max_msg_bytes estimate. A less strict but automated alternative would be to route failed batches to a dead letter queue using a fallback broker, but this would allow subsequent batches to be delivered in the meantime whilst those failed batches are dealt with.

Troubleshooting

• I’m seeing logs that report Failed to connect to kafka: kafka: client has run out of available brokers to talk to (Is your cluster reachable?), but the brokers are definitely reachable.

Unfortunately this error message will appear for a wide range of connection problems even when the broker endpoint can be reached. Double check your authentication configuration and also ensure that you have enabled TLS if applicable.

Performance

This output benefits from sending multiple messages in flight in parallel for improved performance. You can tune the max number of in flight messages (or message batches) with the field max_in_flight.

This output benefits from sending messages as a batch for improved performance. Batches can be formed at both the input and output level.

Fields

addresses

A list of broker addresses to connect to. If an item of the list contains commas it will be expanded into multiple addresses.

Type: array

# Examples

addresses:
  - localhost:9092

addresses:
  - localhost:9041,localhost:9042

addresses:
  - localhost:9041
  - localhost:9042

tls

Custom TLS settings can be used to override system defaults.

Type: object

tls.enabled

Whether custom TLS settings are enabled.

Type: bool
Default: false

tls.skip_cert_verify

Whether to skip server side certificate verification.

Type: bool
Default: false

tls.enable_renegotiation

Whether to allow the remote server to repeatedly request renegotiation. Enable this option if you’re seeing the error message local error: tls: no renegotiation.

Type: bool
Default: false

tls.root_cas

An optional root certificate authority to use. This is a string, representing a certificate chain from the parent trusted root certificate, to possible intermediate signing certificates, to the host certificate.

Type: string
Default: ""

# Examples

root_cas: |-
  -----BEGIN CERTIFICATE-----
  ...
  -----END CERTIFICATE-----  

tls.root_cas_file

An optional path of a root certificate authority file to use. This is a file, often with a .pem extension, containing a certificate chain from the parent trusted root certificate, to possible intermediate signing certificates, to the host certificate.

Type: string
Default: ""

# Examples

root_cas_file: ./root_cas.pem

tls.client_certs

A list of client certificates to use. For each certificate either the fields cert and key, or cert_file and key_file should be specified, but not both.

Type: array
Default: []

# Examples

client_certs:
  - cert: foo
    key: bar

client_certs:
  - cert_file: ./example.pem
    key_file: ./example.key

tls.client_certs[].cert

A plain text certificate to use.

Type: string
Default: ""

tls.client_certs[].key

A plain text certificate key to use.

Type: string
Default: ""

tls.client_certs[].cert_file

The path of a certificate to use.

Type: string
Default: ""

tls.client_certs[].key_file

The path of a certificate key to use.

Type: string
Default: ""

tls.client_certs[].password

A plain text password for when the private key is password encrypted in PKCS#1 or PKCS#8 format. The obsolete pbeWithMD5AndDES-CBC algorithm is not supported for the PKCS#8 format. Warning: Since it does not authenticate the ciphertext, it is vulnerable to padding oracle attacks that can let an attacker recover the plaintext.

Type: string
Default: ""

# Example

password: foo

sasl

Enables SASL authentication.

Type: object

sasl.mechanism

The SASL authentication mechanism, if left empty SASL authentication is not used.

Type: string
Default: "none"

sasl.user

A PLAIN username. It is recommended that you use environment variables to populate this field.

Type: string
Default: ""

# Examples

user: ${USER}

sasl.password

A PLAIN password. It is recommended that you use environment variables to populate this field.

Type: string
Default: ""

# Examples

password: ${PASSWORD}

sasl.access_token

A static OAUTHBEARER access token

Type: string
Default: ""

sasl.token_cache

Instead of using a static access_token allows you to query a cache resource to fetch OAUTHBEARER tokens from

Type: string
Default: ""

sasl.token_key

Required when using a token_cache, the key to query the cache with for tokens.

Type: string
Default: ""

topic

The topic to publish messages to.

Type: string

client_id

An identifier for the client connection.

Type: string
Default: "tyk"

target_version

The version of the Kafka protocol to use. This limits the capabilities used by the client and should ideally match the version of your brokers. Defaults to the oldest supported stable version.

Type: string

# Examples

target_version: 2.1.0

target_version: 3.1.0

rack_id

A rack identifier for this client.

Type: string
Default: ""

key

The key to publish messages with.

Type: string
Default: ""

partitioner

The partitioning algorithm to use.

Type: string
Default: "fnv1a_hash"
Options: fnv1a_hash, murmur2_hash, random, round_robin, manual.

partition

The manually-specified partition to publish messages to, relevant only when the field partitioner is set to manual. Must be able to parse as a 32-bit integer.

Type: string
Default: ""

custom_topic_creation

If enabled, topics will be created with the specified number of partitions and replication factor if they do not already exist.

Type: object

custom_topic_creation.enabled

Whether to enable custom topic creation.

Type: bool
Default: false

custom_topic_creation.partitions

The number of partitions to create for new topics. Leave at -1 to use the broker configured default. Must be >= 1.

Type: int
Default: -1

custom_topic_creation.replication_factor

The replication factor to use for new topics. Leave at -1 to use the broker configured default. Must be an odd number, and less then or equal to the number of brokers.

Type: int
Default: -1

compression

The compression algorithm to use.

Type: string
Default: "none"
Options: none, snappy, lz4, gzip, zstd.

static_headers

An optional map of static headers that should be added to messages in addition to metadata.

Type: object

# Examples

static_headers:
  first-static-header: value-1
  second-static-header: value-2

metadata

Specify criteria for which metadata values are sent with messages as headers.

Type: object

metadata.exclude_prefixes

Provide a list of explicit metadata key prefixes to be excluded when adding metadata to sent messages.

Type: array
Default: []

max_in_flight

The maximum number of messages to have in flight at a given time. Increase this to improve throughput.

Type: int
Default: 64

idempotent_write

Enable the idempotent write producer option. This requires the IDEMPOTENT_WRITE permission on CLUSTER and can be disabled if this permission is not available.

Type: bool
Default: false

ack_replicas

Ensure that messages have been copied across all replicas before acknowledging receipt.

Type: bool
Default: false

max_msg_bytes

The maximum size in bytes of messages sent to the target topic.

Type: int
Default: 1000000

timeout

The maximum period of time to wait for message sends before abandoning the request and retrying.

Type: string
Default: "5s"

retry_as_batch

When enabled forces an entire batch of messages to be retried if any individual message fails on a send, otherwise only the individual messages that failed are retried. Disabling this helps to reduce message duplicates during intermittent errors, but also makes it impossible to guarantee strict ordering of messages.

Type: bool
Default: false

batching

Allows you to configure a batching policy.

Type: object

# Examples

batching:
  byte_size: 5000
  count: 0
  period: 1s

batching:
  count: 10
  period: 1s

batching:
  check: this.contains("END BATCH")
  count: 0
  period: 1m

batching.count

A number of messages at which the batch should be flushed. If 0 disables count based batching.

Type: int
Default: 0

batching.byte_size

An amount of bytes at which the batch should be flushed. If 0 disables size based batching.

Type: int
Default: 0

batching.period

A period in which an incomplete batch should be flushed regardless of its size.

Type: string
Default: ""

# Examples

period: 1s

period: 1m

period: 500ms

batching.processors

A list of processors to apply to a batch as it is flushed. This allows you to aggregate and archive the batch however you see fit. Please note that all resulting messages are flushed as a single batch, therefore splitting the batch into smaller batches using these processors is a no-op.

Type: array

# Examples

processors:
  - archive:
      format: concatenate

processors:
  - archive:
      format: lines

processors:
  - archive:
      format: json_array

max_retries

The maximum number of retries before giving up on the request. If set to zero there is no discrete limit.

Type: int
Default: 0

backoff

Control time intervals between retry attempts.

Type: object

backoff.initial_interval

The initial period to wait between retry attempts.

Type: string
Default: "3s"

# Examples

initial_interval: 50ms

initial_interval: 1s

backoff.max_interval

The maximum period to wait between retry attempts

Type: string
Default: "10s"

# Examples

max_interval: 5s

max_interval: 1m

backoff.max_elapsed_time

The maximum overall period of time to spend on retry attempts before the request is aborted. Setting this value to a zeroed duration (such as 0s) will result in unbounded retries.

Type: string
Default: "30s"

# Examples

max_elapsed_time: 1m

max_elapsed_time: 1h