My Big Data

Thursday, June 11, 2020

Ambari Commands

Ambari Status	/sbin/service ambari-server status
Start Ambari	/sbin/service ambari-server start
Stop Amabri	/sbin/service ambari-server stop
Reset ambari admin pwd	ambari-admin-password-reset
Ambari ldap sync	ambari-server sync-ldap --users users.txt --groups groups.txt
ambari-server --version	Ambari version
ambari-server --hash	Ambari server hash value
Ambari-server backup	To take backup of ambari settings

Decommission a data node in HDP

From Ambari :

Manually :
[root@namenode1 conf]# pwd
/etc/hadoop/conf
[root@namenode1 conf]# ll | grep dfs.exclude
-rw-r--r--. 1 hdfs hadoop 1 Jan 12 15:32 dfs.exclude
[root@namenode1 conf]# cat dfs.exclude

[root@namenode1 conf]# vi dfs.exclude
[root@namenode1 conf]# cat dfs.exclude
edgenode.hdp.cn
[root@namenode1 conf]# su - hdfs
Last login: Sun Jan 12 15:33:08 EST 2020
[hdfs@namenode1 ~]$ hdfs dfsadmin -refreshNodes
Refresh nodes successful

[hdfs@namenode1 ~]$

Friday, February 2, 2018

Disable Resource Manager HA

Stop YARN and Zookeeper services from ambari
From Ambari-server :

                                           /var/lib/ambari-server/resources/scripts/configs.py --user=admin --
   password=admin --host=edgenode.hdp.cn --cluster=hdpdev --
   action=get --config-      type=yarn-site -f yarn-site.json

From the below change the first property to value "false" and remove the other properties from yarn-site.json

a) "yarn.resourcemanager.ha.enabled": "false",

b) "yarn.resourcemanager.ha.rm-ids": "rm1,rm2",

c) "yarn.resourcemanager.hostname.rm1": "datanode1.hdp.cn",

d) "yarn.resourcemanager.hostname.rm2": "edgenode.hdp.cn",

e) "yarn.resourcemanager.webapp.address.rm1": "datanode1.hdp.cn:8088",

f) "yarn.resourcemanager.webapp.address.rm2": "edgenode.hdp.cn:8088",

g) "yarn.resourcemanager.webapp.https.address.rm1": "datanode1.hdp.cn:8090",

h) "yarn.resourcemanager.webapp.https.address.rm2": "edgenode.hdp.cn:8090",

i) "yarn.resourcemanager.cluster-id": "yarn-cluster",

j) "yarn.resourcemanager.ha.automatic-failover.zk-base-path": "/yarn-leader-election",

Set the below properties to existing resource manager

a) "yarn.resourcemanager.hostname":

b) "yarn.resourcemanager.admin.address":

c) "yarn.resourcemanager.webapp.address":

d)"yarn.resourcemanager.resource-tracker.address":

e)"yarn.resourcemanager.scheduler.address":

f)"yarn.resourcemanager.webapp.https.address":

g)"yarn.timeline-service.webapp.address":

h)"yarn.timeline-service.webapp.https.address":

i)"yarn.timeline-service.address":

j)"yarn.log.server.url":

Copy the yarn-site.json file back to the ambari-server and run the below command to set the changes made

/var/lib/ambari-server/resources/scripts/configs.py --user=admin --password=admin -- host=edgenode.hdp.cn --cluster=hdpdev --action=get --config-type=yarn-site

Delete the Resource Manager host comonent

   curl --user admin:admin -i -H "X-Requested-By: ambari" -X DELETE
  http://edgenode.hdp.cn:8080/api/v1/clusters/hdpdev/hosts/edgenode.hdp.cn/
host_components/RESOURCEMANAGER

Start zookeeper service from Ambari.

On zookeeper clients run the below command to change the znode permissions

/usr/hdp/current/zookeeper-client/bin/zkCli.sh getAcl /rmstore/ZKRMStateRoot

/usr/hdp/current/zookeeper-client/bin/zkCli.sh setAcl /rmstore/ZKRMStateRoot world:anyone:rwcda

From Ambari UI, restart zookeeper service and YARN service

Tuesday, January 16, 2018

Yarn Capacity Scheduler

1. /etc/hadoop/conf/yarn-site.xml yarn.resourcemanager.scheduler.class=org.apache.hadoop.yarn.server.resourcemanager.scheduler.capacity.CapacityScheduler

2. Capacity Scheduler settings (/etc/hadoop/conf/capacity-scheduler.xml on resource manager host)

a) Queues are fundamental unit of scheduling in yarn

yarn.scheduler.capacity.root.queues = support,engineering,marketing (parent queues)

yarn.scheduler.capacity.support.queues=training,services (leaf queues)

yarn.scheduler.capacity.engineering=development,qa (leaf queues)

yarn.scheduler.capacity.marketing=sales,advertising (leaf queues)

b) Capacity Scheduler ACL's

Queue users:

yarn.scheduler.capacity.root.acl_submit_application= (asterisk == all users and groups & empty space == deny all users and groups)

yarn.scheduler.capacity.support.acl_submit_applications= sherlock,pacioli,njella hg_group,cd_group

yarn.scheduler.capacity.engineering.acl_submit_applications=user1,user2,user3 group1,group2,group3

yarn.scheduler.capacity.marketing.acl_submit_applications= user4,user2,user5, group4,group5,group6

Queue Administers

yarn.scheduler.capacity.root.acl_administer_queue=

yarn.scheduler.capacity.support.acl_administer_queue= support_admin_group

yarn.scheduler.capacity.engineering.acl_administer_queue=engineering_admin_group

yarn.scheduler.capacity.marketing.acl_administer_queue=engineering_Admin_grp

c) Queue Mappings:

Setting Up Queue Mappings

yarn.scheduler.capacity.queue-mappings=u:njella:engineering,u:chitra:marketing,g:engineering_group:engineering

If group name is same as queue name then

yarn.scheduler.capacity.queue-mappings=u:%user:%primary_group

If user name is same as queue name then

yarn.scheduler.capacity.queue-mappings=u:%user:%user

Enable queue mapping override:

yarn.scheduler.capacity.queue-mappings-override.enable=true

Example:

<name>yarn.scheduler.capacity.queue-mappings</name>

<value>u:njella:engineering, g:webadmins:weblog</value>

</proprty>

If user "njella" explicitly submits an application to the marketing queue, the default queue assignment of engineering is overriden and the application is submitted to the marketing queue.

Thursday, January 4, 2018

Issues

1. Ambari-agent

Sol : Change verify=platform_default to verify=disable in /etc/pythoncert-verification.cfg file

sed -i 's/verify=platform_default/verify=disable/' /etc/python/cert-verification.cfg

-------------------------------------------------------------------------------------

Tuesday, October 17, 2017

How to Synchronize the system clock to Network Time Protocol (NTP) in Linux.

1. If you do not have the ntpd service enabled use the below command to install and enable..

a) yum install ntp.
b) systemctl enable ntpd.

2. Modify the /etc/ntp.conf file as shown in the below image. If you are using Cent OS 7, please modify the cat /etc/ntp/step-tickers.

You can also find the preferred ones for your location at http://www.pool.ntp.org/