Hot Standby and Hot Back-Up

  • Thread starter Schaminee, Bart (IndSys, GEFanuc, Nether
  • Start date
S

Thread Starter

Schaminee, Bart (IndSys, GEFanuc, Nether

Hello List,

I experience some confusion:
What is Hot Standby and what is Hot Back-Up.
The issue is: which of both is bumpless.

Background:
in a lot of specs I have read Hot Standby, but the aimed for a Hot Back-Up system. These two are mixed up every time, when we have a request for a high availibility system.

What is your opinion.
Mail to: [email protected]

Bart Schaminee
 
V

Vitor Finkel

Hi, Bart,

I don't think the terminology here is quite well established or standardized. I have heard of Hot-stand-by back-up systems. Most architecturs I've
seen to accomplish that seemed either too complicated, clumsy, expensive, and full of extra hardware capable of hiding possible bugs or failures.

User's experience feedback I received while teaching for many years in a non-vendor PLC course is disapointing. While most users succeed to make the CPU transfer without too much trouble, during tests, so many of them are disappointed with the fact that when a transfer is actually needed, many systems often fail to transfer. So, IMHO, the main issue is not to
determine if it is Bumpless, but if it is reliable enought to justify the expense.

Not to mention that most systems will not allow for a hot-stand-by transfer or redundancy of the communication chanels between CPUs and I/Os, and so are still subject to total failures due to a single component problem, only it is no longer the CPU.

Maybe those comments do not apply to GE-Fanuc + Genious, due to its unique communication and architectures, but I can't really tell.

Vitor Finkel

Vitor Finkel [email protected]
P.O. Box 16061 tel (+55) 21 285-5641
22222.970 Rio de Janeiro Brazil fax (+55) 21 205-3339
 
R
I think it amounts to two terms meaning essentially the same thing.

I do have a moderately humurous story about this though. One time I saw a control that had an extra PLC2/30 in it. It was powered up but not connected to any I/O. I asked an electrician what it was for. he said it was their "hot backup" PLC. I guess he was right - it was powered so it was hot, and it was a spare, or backup PLC. they had the program loaded in the spare PLC and all thery had to do if the main PLC died was turn it off, move the I/O cable from the main to the spare and turn power back on.
 
Regardless of whether it is called hot stand-by or hot back-up, when you are looking for high availability, you need to have redundancy in both the processor AND the I/O, not just in the processor. Most PLC systems I have seen do not have I/O redundancy as a standard offering. Due to the connection to the "real world", I/O modules have the greater opportunity to fail. The ability to bumplessly place a "hot spare" on line and then to identify, remove and replace the failed module without interrupting the process seems to be critical to providing high availability.

Given the above, the Triconex Triple Mode Redundant systems (TMR) fill the bill quite nicely, but not without emptying your checkbook at the same time. You must weigh the cost of a shutdown due to module failure against the added cost of a TMR system to justify your selection.

Good luck with your search.
 
Top