support - Cloudera Distribution for Hadoop

Cloudera Distribution for Hadoop for Teradata Administrator Guide

Product
Cloudera Distribution for Hadoop
Release Number
5.11
Published
July 2017
Language
English (United States)
Last Update
2018-04-12
dita:mapPath
lrq1499784642284.ditamap
dita:ditavalPath
cdh_Cloudera.ditaval
dita:id
ndz1467245415037
lifecycle
previous
Product Category
Open Source

Purpose

This command allows you to report disk status, configure disks after replacement, install files, and set passwords to match Cloudera Manager and PostgreSQL.

Syntax

hcli [--option] support [checkDisks [--option] ] | [configureDisk [--option]] | [install rpm name | patch name [--option]] | [setpgpass password ] | [rebalance] | [cmpass password ] | [tune_coresidence [--option]]

Parameters

hcli support checkDisks [OPTIONS]
Report status of disk drives in the cluster.
Option Description
--node Reports disks from the local node only (default on datanodes)
--system Reports disks from the entire cluster (default on master nodes)
--json Reports all data in JSON format
--failed Reports only disks with abnormal status
Exit code 0 indicates the command succeeded; exit code 1 if the report failed.
hcli support configureDisk [--check]
Configure, format, and prepare data node disk configurations for use after a disk drive replacement.
  • Create vdisks on RAID controller
  • Create filesystem and disk label if not present
  • Update mount options in /etc/fstab
  • Mount data drive filesystems
All Hadoop services on the node must be stopped with Cloudera Manager before using this command. Services can be started after the command completes with no errors.
Use the --check option to check that the data node disks are correctly configured and labeled. No configuration changes are made.
Exit code 0 indicates the command succeeded; exit code 1 indicates the number of nodes on which the command failed.
hcli support install --rpm name | --patch name [--master | --data | --edge | --all | --dry-run ]
Install RPMs and patches on nodes in the cluster. Either the RPM and name or the patch and name are required.
  • -r | --rpm name Installs the specified RPM file.
  • -p | --patch name Installs the specified patch file.
Option Description
-m | --master Installs on master nodes
-d | --data Installs on data nodes
-e | --edge Installs on edge nodes
-a | --all Installs on all nodes
-n | --dry-run Tests the installation but does not make changes
Exit code 0 indicates the command succeeded; exit code 1 indicates the command failed.
hcli support install --rpm /tmp/new-package-file.rpm --master --edge
hcli support manager_config [OPTION]
Modify various Cloudera Manager related settings.
Option Description
-p | --port port number Update the port number used to connect to Cloudera Manager.
-h | --host= host fqdn Update the hostname used to connect to Cloudera Manager.
-p | --password= admin password Update the Cloudera Manager administrator password.
When using the --password option, use the --nolog option to avoid recording passwords in debugging logs.
Exit code 0 indicates the Cloudera Manager setting updated successfully; exit code 1 indicates the setting could not be updated.
hcli --nolog support manager_config --password=admin password
hcli support rebalance
Adjust parameters to enable high-priority HDFS rebalancing after expanding or contracting a cluster.
Waits until rebalance is complete before exiting.
Exit code 0 indicates the rebalance succeeded; exit code 1 if the HDFS could not be rebalanced.
hcli [--nolog] support setpgpass password
Change the password on postgres and update postgres/.pgpass files on the postgres nodes.
Use the --nolog option to avoid recording passwords in debugging logs.
Exit code 0 indicates the command succeeded; exit code 1 indicates the command failed.
hcli --nolog support setpgpass pswd
hcli support tune_coresidence [--node | --help]
Tune data nodes in the cluster for coresidence configuration.
Option Description
-n | --node FQDN Specifies the fully-qualified domain name of a single node that is tuned to the coresidence configuration.
-n FQDN-NODE1 -n FQDN-NODE2 Specifies the fully-qualified domain names of all nodes that are tuned to the coresidence configuration.
-h | --help Shows help for this command.
Exit code 0 indicates that the tuning for coresidence configuration has completed; exit code 1 indicates an exception has occurred.
hcli support tune_coresidence
hcli support tune_coresidence --node <FQDN> [-h|--help]
hcli support tune_coresidence -n <FQDN-NODE1> -n <FQDN-NODE2> [-h|--help]