Super Server for Large Mining Automation Project

I’ve been in the process automation business for a long time and met many really great folks. One of them, Fluor Canada‘s Steve Priest, I got to know long ago from his days with Emerson local business partner, Norpac Controls.

Steve is now with Fluor Canada and was working with Paul Yanchus with Emerson local business partner Atlantic Controls in Newfoundland, Canada. Steve is part of a very large mining project that is using a DeltaV system for the automation. The project team has 15 people working simultaneously on the control system configuration.

The control system configuration includes around 10,000 control modules with approximately 20,000 device signal tags (DSTs). For this large of a concurrent engineering effort, the project team was advised that as the database grew, that they would have to split it or subdivide the system into zones.

Because the design included the use of class-based modules, the project team wanted to keep the configuration effort together. They initially fine-tuned the existing server, but the performance could not keep up with the number of engineers working concurrently.

Steve described how the project team worked with Dell and combined their expertise with that of the project team’s knowledge of the application to specify the “Super Server”. Tongue firmly planted in cheek, Steve noted that this solution was with a Dell computer as opposed to a fantastic Newfie waiter!

This super server included the fastest processor in the Dell catalog and it was configured with 6 solid-state drives (SSDs) and 2-15k SCSI hard drives. The DeltaV system information was located on 2 SSDs in a RAID 1 configuration. The control system data was configured on 4 SSDs in a RAID 10 configuration, and the project documents on the 2 SCSI drives in a RAID 0 configuration.

The move to the super server gave quite a performance boost to the project team members. Steve noted that the tasks of saving modules dropped 50% and some exports times were reduced by 75%. It’s allowed the team to work flat out to keep pace with the project schedules. And being together in one project database, the super server keeps the whole project team in sync.

The DeltaV Product Engineering team has been in contact with Steve and the project team to investigate certifying the super server configuration as a fully-tested configuration.

Thanks for sharing the story, Steve and I hope the rest of the project goes well. Hopefully, you’ll encounter a few super servers in the fine dining establishments in Newfoundland, too!

MP3 | iTunes

Audio clip: Adobe Flash Player (version 9 or above) is required to play this audio clip. Download the latest version here. You also need to have JavaScript enabled in your browser.

3 comments

  1. That’s an
    example of a brute force method overcoming faults with DeltaV, in today’s world
    of multi-core processor servers, DeltaV is still a single threaded application,
    and the only way to improve its performance is brute force horsepower.

    We run multiple users connected to our standard dual quad-core (16 CPU Total) Pro+’s and we only see 5% CPU
    activity. That is the reason it is slow, it doesn’t make use of multiple
    processors.

    I’m not
    sure where in lies the problem, either the underlying Objectivity engine or DeltaV’s hawkservice?

    It would
    be nice if product engineering optimized their software to utilize the powerful
    hardware which is supplied as standard with DeltaV, rather than forcing additional
    engineering costs on a project.

Leave a Reply