[ https://issues.jboss.org/browse/TEIIDDES-2453?page=com.atlassian.jira.plu... ] Corey Barrett commented on TEIIDDES-2453: ----------------------------------------- My understanding is that there are two patterns. 1. Direct Consumption of Hadoop: "computer all logged into the kerberos domain." We have seen this working this way in our setup of local BI application with ODBC connections. No logins / passwords entered by users. 2. Consumers of Hadoop through BI Services platforms. BI Servers front the database access permissions and provide shared Web Interface to solution: I believe this is going to be through a keytab file from what I am told. -- This message was sent by Atlassian JIRA (v6.3.11#6341)