Most written before, this is a complete environment to build the basis Hadoop2.6.1, on a three-node.
(Built on Hadoop environment, refer to: https://www.cnblogs.com/SysoCjs/p/10835793.html )
Description:
#master, showing the operation on the master node;
# Master, # slave1, # slave2, showing three nodes should operate on;
First, the preparatory work to the official website to download the appropriate flink the tgz archive:
http://mirror.bit.edu.cn/apache/flink/
Here is the download flink-1.6.4-bin-hadoop26-scala_2.11.tgz
Then use scp command (Git Base or other tools can also be) distributed to the master remote machine CentOS7 specified on the directory.
Second, unzip #master
cd /usr/local/src/
tar -zxvf flink-1.6.4-bin-hadoop26-scala_2.11.tgz
Third, modify the configuration
#master
cd ./flink-1.6.4/conf/
1, configure remote address JobManager
vim flink-conf.yaml
Modify the content:
jobmanager.rpc.address: master
2, registered master
vim masters
Modify the contents of this with flink-conf-yarm of rest.port configuration to be the same:
master:8081
3, registration slave
vim slaves
Modify the content:
slave1
slave2
Fourth, the distribution of documents
#master
scp -r /usr/local/src/flink-1.6.4 root@slave1:/usr/local/src/
scp -r /usr/local/src/flink-1.6.4 root@slave2:/usr/local/src/
V. verification
#master
1, start the cluster
cd /usr/local/src/flink-1.6.4/bin
./start-cluster.sh
master node:
slave node:
2, monitoring page
http://192.168.112.10:8081