000034067 - "No CLDBs defined for cluster" error occurs on Warehouse nodes in RSA Security Analytics 10.x

Document created by RSA Customer Support Employee on Oct 4, 2016•Last modified by RSA Customer Support on Jul 9, 2019
Version 4Show Document
  • Comment • 0
  • View in full screen mode

Article Content

Article Number000034067
Applies To
RSA Product Set: RSA Netwitness Logs & Network
   RSA Product/Service Type: MapR Warehouse (SAW)
   RSA Version/Condition: 10.x
   Platform: CentOS
   O/S Version: EL6
IssueCLDBs is not defined for cluster in one of the saw nodes and the following error occurs:

a. Open a SSH session to one of the SAW nodes and execute below commands:

#service mapr-warden stop 
#service mapr-warden start




User-added image

b. Notice the same error occurs when trying to access a file system

#hadoop fs -ls /





 
Cause
One possible cause of the issue is due to a mismatch of gid and uid for the mapr user and mapr group. These ids need to be consistent across all the warehouse nodes. Below is an example of a mismatch of the uid and gid in node 3 compared with the rest of the nodes.

[root@saw-node3 ~]# id mapr
uid=500(mapr) gid=500(mapr) groups=500(mapr),501(shadow)

[root@saw-node2 ~]# id mapr
uid=2147483632(mapr) gid=2147483632(mapr) groups=2147483632(mapr),500(shadow)

[root@saw-node1 conf]# id mapr
uid=2147483632(mapr) gid=2147483632(mapr) groups=2147483632(mapr),500(shadow)



 
ResolutionBelow are the steps to change uid and gid in Warehouse nodes. 
  1. Open an SSH session to a node which is showing different IDs.
  2. Stop the following services on that node with these commands:

    #service mapr-zookeeper stop
    #service mapr-warden stop
    #stop hive2

  3. Execute the following commands to change the uid and gid:

    #usermod -u 2147483632 mapr
    #groupmod -g 2147483632 mapr

  4. Run the command below to make sure that the ids (gid and uid) are the same on all three nodes:

    # id mapr

  5. Restart the services:

    #service mapr-zookeeper start
    #service mapr-warden start
    #start hive2

  6. Make sure that the file system is accessible with this command:

    #hadoop fs -ls /

Attachments

    Outcomes