Cloudera manager platform fault repair record

Background introduction ：
do not know why ssh Not on the server , It's not easy to connect today , But in the visualization platform yarn The interface is shown as follows ↓

Open the cluster's hdfs

Click... In the status summary datanode

Click in , And then click on this


Click host - All hosts
Then click on the role in the page

Click on each character , Until we find the following , Then click restart to restart all four services

problem 2

There is still a problem , Inside , Rebooted , Report red again , Say there is a clock deviation

So I found the server with the clock deviation
then ssh To this server
then

systemctl restart ntpd # Restart the service 
systemctl status ntpd # View service status , If something doesn't report an error, it means it's normal