pump is a daemon that receives realtime binlog from tidb-server and writes in sequential disk files synchronously.
Usage of pump:
-L string
log level: debug, info, warn, error, fatal (default "info")
-V print pump version info
-addr string
addr(i.e. 'host:port') to listen on for client traffic (default "127.0.0.1:8250")
-advertise-addr string
addr(i.e. 'host:port') to advertise to the public
-config string
path to the pump configuration file
-data-dir string
the path to store binlog data
-gc int
recycle binlog files older than gc days, zero means never recycle (default 7)
-heartbeat-interval int
number of seconds between heartbeat ticks (default 2)
-kafka-addrs string
a comma separated list of the kafka broker endpoints (default "127.0.0.1:9092")
-log-file string
log file path
-log-rotate string
log file rotate type, hour/day
-max-message-size int
max msg size producer produce into kafka (default 1073741824)
-metrics-addr string
prometheus pushgateway address, leaves it empty will disable prometheus push
-metrics-interval int
prometheus client push interval in second, set "0" to disable prometheus push (default 15)
-node-id string
the ID of pump node; if not specify, we will generate one from hostname and the listening port
-pd-urls string
a comma separated list of the PD endpoints (default "http://127.0.0.1:2379")
-zookeeper-addrs string
a comma separated list of the zookeeper broker endpoints
./bin/pump -socket unix:///tmp/pump.sock \
-pd-urls http://127.0.0.1:2379 \
-data-dir ./data.pump
or use configuration file
./bin/pump -config ./conf/pump.toml
You should deployment pump server for each TiDB node in the cluster, pump can serve tidb-servers across different clusters.