Cdh5 delete data node

This article describes how to CDH Hadoop Dynamically delete data nodes in the cluster . Before dynamically deleting data nodes, we already have a set of CDH Cluster environment . The following steps are to delete a data node ,
notes ： Delete nodes online. Try to delete nodes one by one , After deleting a node, perform the necessary checks . If you delete more nodes at a time , Problems that may result in the loss of data blocks .

1. Select the node to delete
Select menu bar " host ", Check the node to be deleted

2. Stop all roles on the host
After checking the node , from " Selected actions " In the drop-down box , choice " Stop the role on the host "
After completion, you can see that all characters on the node are in the stopped state

3. To revoke authority
After checking the node , from " Selected actions " In the drop-down box , choice "Begin Maintenance(Suppress Alerts/Decommission)"
After completion, you can see that the status changes to " Authorization has been revoked "



4. Remove... From the cluster
After checking the node , from " Selected actions " In the drop-down box , choice " Remove... From the cluster "
complete Then you can see this node in the cluster , but CM You can still see this node in



5. from CM Delete in
After checking the node , from " Selected actions " In the drop-down box , choice "Remove From Cloudera Manager"
When it's done CM The page no longer sees this node


6. Redeploy the client


7. Verify if the block is missing
After deleting a node , What we are most concerned about is whether the data is lost , Can pass hadoop fsck / Command check , If the output is as follows, it means that no data block is lost .

....................................................................Status: HEALTHY
 Total size:    10356256081639 B (Total open files size: 996754639 B)
 Total dirs:    280130
 Total files:   453468
 Total symlinks:                0 (Files currently being written: 548)
 Total blocks (validated):      458431 (avg. block size 22590653 B) (Total open file blocks (not validated): 431)
 Minimally replicated blocks:   458431 (100.0 %)
 Over-replicated blocks:        0 (0.0 %)
 Under-replicated blocks:       0 (0.0 %)
 Mis-replicated blocks:         0 (0.0 %)
 Default replication factor:    3
 Average block replication:     3.0
 Corrupt blocks:                0
 Missing replicas:              0 (0.0 %)
 Number of data-nodes:          18
 Number of racks:               1
FSCK ended at Wed Jun 10 14:12:04 CST 2020 in 5169 milliseconds