# TiDB Binlog Drainer Configurations in Kubernetes

This document introduces the configuration parameters for a TiDB Binlog drainer in Kubernetes.

## Configuration parameters

The following table contains all configuration parameters available for the tidb-drainer chart. Refer to Use Helm to learn how to configure these parameters.

Parameter Description Default Value
clusterName The name of the source TiDB cluster demo
clusterVersion The version of the source TiDB cluster v3.0.1
baseImage The base image of TiDB Binlog pingcap/tidb-binlog
imagePullPolicy The pulling policy of the image IfNotPresent
logLevel The log level of the drainer process info
storageClassName storageClass used by the drainer. storageClassName refers to a type of storage provided by the Kubernetes cluster, which might map to a level of service quality, a backup policy, or to any policy determined by the cluster administrator. Detailed reference: storage-classes local-storage
storage The storage limit of the drainer Pod. Note that you should set a larger size if db-type is set to pb 10Gi
disableDetect Determines whether to disable casualty detection false
initialCommitTs Used to initialize a checkpoint if the drainer does not have one 0
config The configuration file passed to the drainer. Detailed reference: drainer.toml (see below)
resources The resource limits and requests of the drainer Pod {}
nodeSelector Ensures that the drainer Pod is only scheduled to the node with the specific key-value pair as the label. Detailed reference: nodeselector {}
tolerations Applies to drainer Pods, allowing the Pods to be scheduled to the nodes with specified taints. Detailed reference: taint-and-toleration {}
affinity Defines scheduling policies and preferences of the drainer Pod. Detailed reference:affinity-and-anti-affinity {}

The default value of config is:

detect-interval = 10
compressor = ""
[syncer]
worker-count = 16
disable-dispatch = false
ignore-schemas = "INFORMATION_SCHEMA,PERFORMANCE_SCHEMA,mysql"
safe-mode = false
txn-batch = 20
db-type = "file"
[syncer.to]
dir = "/data/pb"