Emerson Exchange 365

• Not Answered

DeltaV Redundancy

Dear Experts,

I have a question regarding DeltaV Redundancy.

I have an existing system with DeltaV V5 and I need to upgrade the system which has 1 proplus station; however, in my upgraded system, I need to have full redundancy in all aspects of the system (i.e. Server redundancy, controller redundancy, network redundancy, etc.). I have been able to find some info regarding all types except server redundancy.

The scenario I am looking for is that I should have two Proplus stations, with both of their databases for alarms and histories synced and in case of failure of one of the servers the other will take over and all the clients that are connected to the first server should be connected to the second server; and once the failed server is back in service it should sync with the first server.

Can this be done? are there any documentations regarding the same that I can present to my client?

Thanks

mohammad.abusido

• 11 Sep 2017 7:43 AM
• Cancel

6 Replies

• Brian Atkinson

  

  • 11 Sep 2017 3:32 PM

  Your best (and I believe only) bet for ProPlus redundancy is to run your ProPlus as a Virtual Machine in DeltaV Virtual Studio and take advantage of the failover capabilities in the virtualization solution.

  • Cancel
• Michael Moody

  

  • 11 Sep 2017 3:58 PM

  Can you clarify your requirements for redundancy, especially for two Proplus stations with a synchronized database? Do you have Batch Servers?
  
  Network redundancy is standard on a production system. Controller redundancy is very common (redundant power supply and cpu).
  For a ProPlus, it is not normally needed 100% as Operator Stations. Historians, and Batch Servers can function without ProPlus. There are options for Redundant Batch Servers.
  Virtual servers with appropriately configured failover should be adequate for the ProPlus.
  I would ask Emerson directly for input on the possibilities if Batch redundancy / Historian redundancy is needed.
  
  In any scenario, your Emerson contact should be involved in designed the architecture of such a system.
  
  Michael

  • Cancel
• Andrew Kennard

  

  • 11 Sep 2017 11:39 PM

  I would be interested in understanding your reasons for wanting to have a redundant Proplus. The Proplus is not a 'server', its really just and engineering station with extra features; the plant can run forever without it.
  It does authorise licensing, meaning config changes cannot be made until its back online. It also keeps an offline copy of the total database, but this itself can be thought of as a backup to the realtime database actually running in the controllers and workstations. Trending etc is not normally kept on the Proplus but normally an (optionally redundant) App station. I am not aware of any system that has Proplus redundancy.

  • Cancel
• Youssef.El-Bahtimy

  

  • 12 Sep 2017 12:17 AM

  In reply to Andrew Kennard:

  I don't necessarily agree that the plant can run forever without a proplus, at least not practically. You can't open any database application without the database, so no control studio online for troubleshooting, no version control, no adding or removing users, and in the event it is your only dc (not a recommended practice), no uncached logins. Further more, to backup the database you have to disconnect clients, which a 'redundant' set up could eliminate the need for. I think the virtual studio paradigm solves a lot of these problems, by increasing availability, but a fault in a critical process on one virtual server still brings down the server.

  • Cancel
• Andrew Kennard

  

  • 12 Sep 2017 12:26 AM

  In reply to Youssef.El-Bahtimy:

  I agree that running 'forever' is only theoretical and not practical and I would never recommend it. But my point was that a failed Proplus will not trigger a plant shutdown like other parts of the architecture so the need for redundancy is not critical (or even required)

  • Cancel
• mohammad.abusido

  

  • 12 Sep 2017 6:34 AM

  Thanks for all the answers. I've had the idea that the Proplus station acts as a server for the operator stations in a setup like server - client architecture; now it's clear.

  • Cancel