Use JSON-based configuration format for Hadoop datanodes to create a whitelist and control put these in normal, decommissioned or maintenance state.
This functionality allows using maintenance mode which is useful for short-term operations like system upgrade or reboot operation. This is an advantage over decommissioning datanode which would be useful for long-term operations.
Display HDFS report.
$ hdfs dfsadmin -report
Configured Capacity: 63010750464 (58.68 GB)
Present Capacity: 52226306048 (48.64 GB)
DFS Remaining: 43579908096 (40.59 GB)
DFS Used: 8646397952 (8.05 GB)
DFS Used%: 16.56%
Replicated Blocks:
Under replicated blocks: 14
Blocks with corrupt replicas: 0
Missing blocks: 0
Missing blocks (with replication factor 1): 0
Low redundancy blocks with highest priority to recover: 0
Pending deletion blocks: 0
Erasure Coded Block Groups:
Low redundancy block groups: 0
Block groups with corrupt internal blocks: 0
Missing block groups: 0
Low redundancy blocks with highest priority to recover: 0
Pending deletion blocks: 0
-------------------------------------------------
Live datanodes (3):
Name: 192.168.8.173:9866 (datanode1.example.org)
Hostname: datanode1.example.org
Decommission Status : Normal
Configured Capacity: 21003583488 (19.56 GB)
DFS Used: 2378018816 (2.21 GB)
Non DFS Used: 2480680960 (2.31 GB)
DFS Remaining: 15054364672 (14.02 GB)
DFS Used%: 11.32%
DFS Remaining%: 71.68%
Configured Cache Capacity: 0 (0 B)
Cache Used: 0 (0 B)
Cache Remaining: 0 (0 B)
Cache Used%: 100.00%
Cache Remaining%: 0.00%
Xceivers: 1
Last contact: Sat Jul 17 21:51:37 UTC 2021
Last Block Report: Sat Jul 17 21:43:38 UTC 2021
Num of Blocks: 3470
Name: 192.168.8.174:9866 (datanode2.example.org)
Hostname: datanode2.example.org
Decommission Status : Normal
Configured Capacity: 21003583488 (19.56 GB)
DFS Used: 3345268736 (3.12 GB)
Non DFS Used: 2472132608 (2.30 GB)
DFS Remaining: 14095663104 (13.13 GB)
DFS Used%: 15.93%
DFS Remaining%: 67.11%
Configured Cache Capacity: 0 (0 B)
Cache Used: 0 (0 B)
Cache Remaining: 0 (0 B)
Cache Used%: 100.00%
Cache Remaining%: 0.00%
Xceivers: 1
Last contact: Sat Jul 17 21:51:38 UTC 2021
Last Block Report: Sat Jul 17 21:43:45 UTC 2021
Num of Blocks: 10736
Name: 192.168.8.175:9866 (datanode3.example.org)
Hostname: datanode3.example.org
Decommission Status : Normal
Configured Capacity: 21003583488 (19.56 GB)
DFS Used: 2923110400 (2.72 GB)
Non DFS Used: 2560073728 (2.38 GB)
DFS Remaining: 14429880320 (13.44 GB)
DFS Used%: 13.92%
DFS Remaining%: 68.70%
Configured Cache Capacity: 0 (0 B)
Cache Used: 0 (0 B)
Cache Remaining: 0 (0 B)
Cache Used%: 100.00%
Cache Remaining%: 0.00%
Xceivers: 1
Last contact: Sat Jul 17 21:51:39 UTC 2021
Last Block Report: Sat Jul 17 21:43:45 UTC 2021
Num of Blocks: 7598
Create a simplest possible hosts.json file on a namenode that will function as a whitelist.
$ cat <<EOF | sudo -u hadoop tee /opt/hadoop/hadoop-3.2.2/etc/hadoop/hosts.json
[
{
"hostName": "datanode1.example.org"
},
{
"hostName": "datanode2.example.org"
},
{
"hostName": "datanode3.example.org"
}
]
EOF
Define dfs.namenode.hosts.provider.classname and dfs.hosts options inside hdfs-site.xml on a namenode.
$ sudo -u hadoop vim /opt/hadoop/hadoop-3.2.2/etc/hadoop/hdfs-site.xml
<?xml version="1.0" encoding="UTF-8"?>
<?xml-stylesheet type="text/xsl" href="configuration.xsl"?>
<!--
Licensed under the Apache License, Version 2.0 (the "License");
you may not use this file except in compliance with the License.
You may obtain a copy of the License at
http://www.apache.org/licenses/LICENSE-2.0
Unless required by applicable law or agreed to in writing, software
distributed under the License is distributed on an "AS IS" BASIS,
WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
See the License for the specific language governing permissions and
limitations under the License. See accompanying LICENSE file.
-->
<!-- Put site-specific property overrides in this file. -->
<configuration>
<property>
<name>dfs.name.dir</name>
<value>/opt/hadoop/local_data/namenode</value>
</property>
<property>
<name>dfs.namenode.secondary.http-address</name>
<value>https://secondarynamenode.example.org:9870</value>
</property>
<property>
<name>dfs.replication</name>
<value>2</value>
</property>
<property>
<name>dfs.namenode.hosts.provider.classname</name>
<value>org.apache.hadoop.hdfs.server.blockmanagement.CombinedHostFileManager</value>
</property>
<property>
<name>dfs.hosts</name>
<value>/opt/hadoop/hadoop-3.2.2/etc/hadoop/hosts.json</value>
</property>
</configuration>
Restart service on a namenode.
$ sudo systemctl restart hadoop-namenode.service
Inspect datanodes.
$ hdfs dfsadmin -report
Configured Capacity: 63010750464 (58.68 GB)
Present Capacity: 52226273280 (48.64 GB)
DFS Remaining: 43579875328 (40.59 GB)
DFS Used: 8646397952 (8.05 GB)
DFS Used%: 16.56%
Replicated Blocks:
Under replicated blocks: 14
Blocks with corrupt replicas: 0
Missing blocks: 0
Missing blocks (with replication factor 1): 0
Low redundancy blocks with highest priority to recover: 0
Pending deletion blocks: 0
Erasure Coded Block Groups:
Low redundancy block groups: 0
Block groups with corrupt internal blocks: 0
Missing block groups: 0
Low redundancy blocks with highest priority to recover: 0
Pending deletion blocks: 0
-------------------------------------------------
Live datanodes (3):
Name: 192.168.8.173:9866 (datanode1.example.org)
Hostname: datanode1.example.org
Decommission Status : Normal
Configured Capacity: 21003583488 (19.56 GB)
DFS Used: 2378018816 (2.21 GB)
Non DFS Used: 2480693248 (2.31 GB)
DFS Remaining: 15054352384 (14.02 GB)
DFS Used%: 11.32%
DFS Remaining%: 71.68%
Configured Cache Capacity: 0 (0 B)
Cache Used: 0 (0 B)
Cache Remaining: 0 (0 B)
Cache Used%: 100.00%
Cache Remaining%: 0.00%
Xceivers: 1
Last contact: Sat Jul 17 23:19:41 UTC 2021
Last Block Report: Sat Jul 17 23:09:17 UTC 2021
Num of Blocks: 3470
Name: 192.168.8.174:9866 (datanode2.example.org)
Hostname: datanode2.example.org
Decommission Status : Normal
Configured Capacity: 21003583488 (19.56 GB)
DFS Used: 3345268736 (3.12 GB)
Non DFS Used: 2472144896 (2.30 GB)
DFS Remaining: 14095650816 (13.13 GB)
DFS Used%: 15.93%
DFS Remaining%: 67.11%
Configured Cache Capacity: 0 (0 B)
Cache Used: 0 (0 B)
Cache Remaining: 0 (0 B)
Cache Used%: 100.00%
Cache Remaining%: 0.00%
Xceivers: 1
Last contact: Sat Jul 17 23:19:41 UTC 2021
Last Block Report: Sat Jul 17 23:09:17 UTC 2021
Num of Blocks: 10736
Name: 192.168.8.175:9866 (datanode3.example.org)
Hostname: datanode3.example.org
Decommission Status : Normal
Configured Capacity: 21003583488 (19.56 GB)
DFS Used: 2923110400 (2.72 GB)
Non DFS Used: 2560081920 (2.38 GB)
DFS Remaining: 14429872128 (13.44 GB)
DFS Used%: 13.92%
DFS Remaining%: 68.70%
Configured Cache Capacity: 0 (0 B)
Cache Used: 0 (0 B)
Cache Remaining: 0 (0 B)
Cache Used%: 100.00%
Cache Remaining%: 0.00%
Xceivers: 1
Last contact: Sat Jul 17 23:19:41 UTC 2021
Last Block Report: Sat Jul 17 23:09:17 UTC 2021
Num of Blocks: 7598
Put the datanode2.example.org in maintenance mode and datanode3.example.org into decommissioned state.
$ cat << | sudo -u hadoop tee /opt/hadoop/hadoop-3.2.2/etc/hadoop/hosts.json
[
{
"hostName": "datanode1.example.org",
"adminState": "NORMAL"
},
{
"hostName": "datanode2.example.org",
"adminState": "IN_MAINTENANCE"
},
{
"hostName": "datanode3.example.org",
"adminState": "DECOMMISSIONED"
}
]
EOF
Re-read the hosts.json file on a namenode.
$ hdfs dfsadmin -refreshNodes
Refresh nodes successful
Display HDFS report to inspect datanodes.
$ hdfs dfsadmin -report
Configured Capacity: 21003583488 (19.56 GB)
Present Capacity: 17432244957 (16.24 GB)
DFS Remaining: 15054168488 (14.02 GB)
DFS Used: 2378076469 (2.21 GB)
DFS Used%: 13.64%
Replicated Blocks:
Under replicated blocks: 7586
Blocks with corrupt replicas: 0
Missing blocks: 0
Missing blocks (with replication factor 1): 0
Low redundancy blocks with highest priority to recover: 7572
Pending deletion blocks: 0
Erasure Coded Block Groups:
Low redundancy block groups: 0
Block groups with corrupt internal blocks: 0
Missing block groups: 0
Low redundancy blocks with highest priority to recover: 0
Pending deletion blocks: 0
-------------------------------------------------
Live datanodes (3):
Name: 192.168.8.173:9866 (datanode1.example.org)
Hostname: datanode1.example.org
Decommission Status : Normal
Configured Capacity: 21003583488 (19.56 GB)
DFS Used: 2378076469 (2.21 GB)
Non DFS Used: 2480721611 (2.31 GB)
DFS Remaining: 15054168488 (14.02 GB)
DFS Used%: 11.32%
DFS Remaining%: 71.67%
Configured Cache Capacity: 0 (0 B)
Cache Used: 0 (0 B)
Cache Remaining: 0 (0 B)
Cache Used%: 100.00%
Cache Remaining%: 0.00%
Xceivers: 4
Last contact: Sat Jul 17 23:25:50 UTC 2021
Last Block Report: Sat Jul 17 23:09:17 UTC 2021
Num of Blocks: 3478
Name: 192.168.8.174:9866 (datanode2.example.org)
Hostname: datanode2.example.org
Decommission Status : Entering maintenance
Configured Capacity: 21003583488 (19.56 GB)
DFS Used: 3345268736 (3.12 GB)
Non DFS Used: 2472144896 (2.30 GB)
DFS Remaining: 14095650816 (13.13 GB)
DFS Used%: 15.93%
DFS Remaining%: 67.11%
Configured Cache Capacity: 0 (0 B)
Cache Used: 0 (0 B)
Cache Remaining: 0 (0 B)
Cache Used%: 100.00%
Cache Remaining%: 0.00%
Xceivers: 1
Last contact: Sat Jul 17 23:25:50 UTC 2021
Last Block Report: Sat Jul 17 23:09:17 UTC 2021
Num of Blocks: 10736
Name: 192.168.8.175:9866 (datanode3.example.org)
Hostname: datanode3.example.org
Decommission Status : Decommission in progress
Configured Capacity: 21003583488 (19.56 GB)
DFS Used: 2923110400 (2.72 GB)
Non DFS Used: 2560081920 (2.38 GB)
DFS Remaining: 14429872128 (13.44 GB)
DFS Used%: 13.92%
DFS Remaining%: 68.70%
Configured Cache Capacity: 0 (0 B)
Cache Used: 0 (0 B)
Cache Remaining: 0 (0 B)
Cache Used%: 100.00%
Cache Remaining%: 0.00%
Xceivers: 1
Last contact: Sat Jul 17 23:25:50 UTC 2021
Last Block Report: Sat Jul 17 23:09:17 UTC 2021
Num of Blocks: 7598
Decommissioning datanodes (1):
Name: 192.168.8.175:9866 (datanode3.example.org)
Hostname: datanode3.example.org
Decommission Status : Decommission in progress
Configured Capacity: 21003583488 (19.56 GB)
DFS Used: 2923110400 (2.72 GB)
Non DFS Used: 2560081920 (2.38 GB)
DFS Remaining: 14429872128 (13.44 GB)
DFS Used%: 13.92%
DFS Remaining%: 68.70%
Configured Cache Capacity: 0 (0 B)
Cache Used: 0 (0 B)
Cache Remaining: 0 (0 B)
Cache Used%: 100.00%
Cache Remaining%: 0.00%
Xceivers: 1
Last contact: Sat Jul 17 23:25:50 UTC 2021
Last Block Report: Sat Jul 17 23:09:17 UTC 2021
Num of Blocks: 7598
Entering maintenance datanodes (1):
Name: 192.168.8.174:9866 (datanode2.example.org)
Hostname: datanode2.example.org
Decommission Status : Entering maintenance
Configured Capacity: 21003583488 (19.56 GB)
DFS Used: 3345268736 (3.12 GB)
Non DFS Used: 2472144896 (2.30 GB)
DFS Remaining: 14095650816 (13.13 GB)
DFS Used%: 15.93%
DFS Remaining%: 67.11%
Configured Cache Capacity: 0 (0 B)
Cache Used: 0 (0 B)
Cache Remaining: 0 (0 B)
Cache Used%: 100.00%
Cache Remaining%: 0.00%
Xceivers: 1
Last contact: Sat Jul 17 23:25:50 UTC 2021
Last Block Report: Sat Jul 17 23:09:17 UTC 2021
Num of Blocks: 10736
After a while, the state will change.
[...] In maintenance datanodes (1): Name: 192.168.8.174:9866 (datanode2.example.org) Hostname: datanode2.example.org Decommission Status : In maintenance Configured Capacity: 21003583488 (19.56 GB) DFS Used: 3350933504 (3.12 GB) Non DFS Used: 2473369600 (2.30 GB) DFS Remaining: 14088761344 (13.12 GB) DFS Used%: 15.95% DFS Remaining%: 67.08% Configured Cache Capacity: 0 (0 B) Cache Used: 0 (0 B) Cache Remaining: 0 (0 B) Cache Used%: 100.00% Cache Remaining%: 0.00% Xceivers: 1 Last contact: Sun Jul 18 00:14:20 UTC 2021 Last Block Report: Sat Jul 17 23:09:17 UTC 2021 Num of Blocks: 10836 [...]
Additional notes
By default the minimal number of replicas that datanode undergoing maintenance needs to satisfy is set to 1. Alter dfs.namenode.maintenance.replication.min option in hdfs-site.xml configuration file to increase the minimal number of replicas.
You can define when maintenance will expire.
$ date Sun Jul 18 00:01:23 UTC 2021
$ date +%s 1626566485
Calculate maintenance expire time in ms after 1 hour.
$ date +"%s000" --date "+1 hour" 1626570086000
Define it in hosts.json file, remember to refresh nodes.
$ sudo -u hadoop cat /opt/hadoop/hadoop-3.2.2/etc/hadoop/hosts.json
[
{
"hostName": "datanode1.example.org",
"adminState": "NORMAL"
},
{
"hostName": "datanode2.example.org",
"adminState": "IN_MAINTENANCE",
"maintenanceExpireTimeInMS": 1626570086000
},
{
"hostName": "datanode3.example.org",
"adminState": "NORMAL"
}
]
Useful links: