Configure Kerberos

You can specify Kerberos options with any output or input that supports Kerberos, like Elasticsearch.

The following encryption types are supported:

Example output config with Kerberos password based authentication:

output.elasticsearch.hosts: ["http://my-elasticsearch.elastic.co:9200"]
output.elasticsearch.kerberos.auth_type: password
output.elasticsearch.kerberos.username: "elastic"
output.elasticsearch.kerberos.password: "changeme"
output.elasticsearch.kerberos.config_path: "/etc/krb5.conf"
output.elasticsearch.kerberos.realm: "ELASTIC.CO"

The service principal name for the Elasticsearch instance is contructed from these options. Based on this configuration it is going to be HTTP/my-elasticsearch.elastic.co@ELASTIC.CO.

You can specify the following options in the kerberos section of the journalbeat.yml config file:

The enabled setting can be used to enable the kerberos configuration by setting it to false. The default value is true.

There are two options to authenticate with Kerberos KDC: password and keytab.

password expects the principal name and its password. When choosing keytab, you have to specify a princial name and a path to a keytab. The keytab must contain the keys of the selected principal. Otherwise, authentication will fail.

You need to set the path to the krb5.conf, so Journalbeat can find the Kerberos KDC to retrieve a ticket.

Name of the principal used to connect to the output.

If you configured password for auth_type, you have to provide a password for the selected principal.

If you configured keytab for auth_type, you have to provide the path to the keytab of the selected principal.

This option can only be configured for Kafka. It is the name of the Kafka service, usually kafka.

Name of the realm where the output resides.