Puppet Class: profile::analytics::cluster::users

Defined in:
modules/profile/manifests/analytics/cluster/users.pp

Overview

Class profile::analytics::cluster::users

Installs any special system users needed on analytics namenodes or clients. This is used for ensuring that users exist for use in HDFS.

NOTE: Puppet does not manage creation of system user HDFS home directories. you will need to do this manually. To do so, run from any Hadoop node:

sudo -u hdfs hdfs dfs -mkdir /user/$user && \
sudo -u hdfs hdfs dfs -chown $user:$group /user/$user

And optionally:

sudo -u hdfs hdfs dfs -chmod 775 /user/$user


13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
# File 'modules/profile/manifests/analytics/cluster/users.pp', line 13

class profile::analytics::cluster::users {
    # analytics-search user will be used to deploy
    # wikimedia/discovery/analytics into HDFS.
    # The analytics-search-users group will be allowed to
    # sudo -u analytics-search.
    group { 'analytics-search':
        ensure => present,
    }

    user { 'analytics-search':
        ensure => present,
        gid    => 'analytics-search',
        system => true,
    }

    # The analytics user will be used to run any Analytics
    # job running on Yarn/HDFS (as replacement for 'hdfs').
    # T220971
    user { 'analytics':
        ensure => present,
        system => true,
    }

    # The analytics-privatedata user will be used to run
    # cronjobs and similar by users.
    # T238306
    user { 'analytics-privatedata':
        ensure => present,
        system => true,
    }

    # The analytics-product user will be used to run
    # cronjobs and similar by Product Analytics.
    # T255039
    user { 'analytics-product':
        ensure => present,
        system => true,
    }

    # When Kerberos is enabled, indexation jobs will run on workers
    # as user 'druid'.
    class { '::druid::cdh::hadoop::user': }
}