Highlights

Stone Ram Catalog Repair Service gets environment backup and running.

Problem

A large luxury Rug Manufacturing company in India used to have a virtual machine at their NetBackup master server.  A few years back, they installed a new Netabckup 5240 Appliance as the master.  This appliance had been installed using the same name as the original master "OLD-NAME".  The catalog from the virtual master had been recovered to the NetBackup appliance, and the hostname had been updated to a "NEW-NAME" and the "OLD-NAME" was added as an alias in the local hosts file. Internaly in the netbackup EMM it was stull using the OLD-NAME, but externaly host would communicate using the "NEW-NAME".

The customer has sucessfully upgraded from 2.5 to 3,1, however a recent upgrade to 3.1.2 resulted in communication and certification issues.

The customer first engaged with Vertias Technical Support to try to fix the issues, but after it was determined that the issues were due to a previous rename, support were unable to assist.   The customer then requested consulting assistance to resolve the issue as their environment had now been "down" for an extended period of time.

 

Solution

One of our NetBackup catalog specialists, investigated the environment and identified, that there were many inconsistencies between the hostnames being used. 

  • At the operating system (linux based) the hostname was "NEWNAME"
  • Internal to the NBDB the hostname was "OLDNAME"
  • Appliance based web components and clish also used "OLDNAME"
  • NetBackup configuration was using a mixture of both "OLDNAME" and "NEWNAME"
  • MSDP was using a mixture of "OLDNAME" and "NEWNAME"
  • Relational Database Engine was using "OLDNAME"

It was determined that the original rename had missed some important steps, The consultant manualy worked at renaming the various components to make the environment consistent again.

  • Removed the alias from the host file so the underlying issues were not being masked.
  • Investigated all of the comonents of a NetBackup Appliance to determinw where the naming issues were
  • Updated the Relationalal Database Configuration to use the correct name
  • Manipulated and corrected the name in the NBDB so they were using the correct name
  • Corrected all the names in the NetBackup configuration
  • Recovered and corrected all the certificates so clients could continue to communicate with the master.
  • Reconfigured the MSDP to use the correct name.
  • Cleaned up foreign_media entries from the original rename.

 

Impact

The environment was up and running again after 7 hours, allowing  for nightly backups to take place. Continuation of investigation and cleanup of MSDP issue was resolved the following day outside of the backup window.