Fed up with RSViewSE distributed redundant

OkiePC

Lifetime Supporting Member
Join Date
Mar 2005
Location
ENE of Nowhere Oklahoma
Posts
11,791
Our three grinding operations HMI platforms use RSViewSE 4.0 distrubuted for both beef and sausage grinding operations as the HMI.

There are four servers that handle the apps, two servers, two clients, and the HMI logs in via thin clients through citrix. There is yet another server hosting the Factory talk directory and RSSQL for data logging, and a sixth server solely used for domain control and providing a path to the plant LAN. When there is a problem, the whole plant goes down while we scramble to get the system running again. This has happened at least 8 times in the past year and a half (since I started workign here).

Often, it can take a couple of hours to get running again and we rarely know for sure what the cause was. It seems that the redundancy option causes most of our issues, but there have been a couple of times where the system has performed a failover and actually prevented downtime. All in all, the downtime we have expereinced, IMO, has been much greater than what we'd have with 3 standalone PC based HMI systems.

My counterpart has spent countless man hours patching, tweaking, studying and developing the HMI stuff. He has worked with Rockwell and tried all sorts of arrangements related to which servers host which peices of the jigsaw puzzle, but no matter what, there have been problems.

We have decided to go to a standalone system, so that downtime will at least be limited to 1/3 of the plant, and will be more manageable for us controls folks without so much dependence on IT.

Mike (my counterpart and our SE guru) is looking into RSViewSE standalone so that he won't lose all of his development. He frequently says he wishes he'd pushed mgmt to use wonderware or Ifix at the beginning, but now it would be like starting over to do so.

What I am wondering is would it be better to use RSView32, and would it be possible to re-use any of the RSViewSE stuff if we did that?

Also, what is the difference between:
FactoryTalk View Site Edition Station (ie 9701-VWSB100AENE)
FactoryTalk View Machine Edition Station for Windows (ie 9701-VWMR075AENE)

Thanks,
Paul
 
Last edited:
If you were to go to RSView 32 it would be like going to wonderware or some other HMI.

In the past we have also had to switch to SE Standalone from the distrubuted SE. It is a direct import into SE just need to create a standalone application and import your screens, macro's and anything else that is in the project. Only problem I see is if you used different areas inside distributed but you would just have to import all screens into the standalone along with merge the tag databases into one if you did not use direct tag naming.
 
Well, thats the exact reason why we got away from RSView, actually all AB hmis in general. We moved to FactoryPMI/FactorySQL. Using 2 servers, you can have a completely redundant system, and use your existing client machines
 
Yes, I would like to go with another vendor altogether, and we would if it were my decision alone. It is Mike's baby and he has put a boatload of work into it.

He is using direct tag addressing for almost every tag, and is using parameter files extensively. Also there is a considerable amount of VBA involved which sends data to and from an application he refres to as a "web solver". It is an algorithm that calculates ideal batch sizes with parts from up to four different initial blenders in order to obtain the exact fat percentage in the final batches.

It has grown into quite a Frankenstein of a system, and we will be expected to make the transition to standalone systems very quickly. Also, I believe there is a negotiation ongoing between our supplier and Rockwell to give us some sort of allowance when we "trade down" from SE "distributed" to "stations".

The PLCs are all SLC BTW, and there is a controllogix gateway in the mix too.
 
OkiePC said:
Also, what is the difference between:
FactoryTalk View Site Edition Station (ie 9701-VWSB100AENE)
FactoryTalk View Machine Edition Station for Windows (ie 9701-VWMR075AENE)

Thanks,
Paul

Site Edition Station is the old RsView SE that runs on a PC
Machine Edition station allows a PC to behave as PV+

Main differnce I know between SE and ME, the SE allows scripts to run on pushbutton actions, and I believe also allows VBA whereas ME does not
 
Paul,
That situation is out of control - I feel for you. In talking with lots of integrators I've found that RSView SE - especially with redundancy is clunky at best - Rockwell's problem child. This obviously doesn't help you now, but is good to put out to those who are considering it.

Hopefully the stand alone systems don't greatly decrease your operator capabilities. They will certainly hinder scalability at some point, but if the system meets your requirements, more power to you. I'm a big fan of simplicity.

For your case, virtualization likely would have made some things easier. 2 physical servers running those 8 virtual machines (with VMotion) provides several advantages:
1. One physical server could go down and the other would run everything automatically.
2. Reboots are quicker and easier.
3. Upgrades are smoother because you can backup entire "images" and more easily fix problems. Do not try to move your existing problematic system over - that would be introducing complexity where you don't need it.

That said, FactoryPMI is easily as capable as RSView SE and has a redundancy system that actually works and is simple to use. The key is that it's architecturally simple. But you'd have to start from scratch just like with Wonderware or iFix.

OkiePC said:
Our three grinding operations HMI platforms use RSViewSE 4.0 distrubuted for both beef and sausage grinding operations as the HMI.

There are four servers that handle the apps, two servers, two clients, and the HMI logs in via thin clients through citrix. There is yet another server hosting the Factory talk directory and RSSQL for data logging, and a sixth server solely used for domain control and providing a path to the plant LAN. When there is a problem, the whole plant goes down while we scramble to get the system running again. This has happened at least 8 times in the past year and a half (since I started workign here).

Often, it can take a couple of hours to get running again and we rarely know for sure what the cause was. It seems that the redundancy option causes most of our issues, but there have been a couple of times where the system has performed a failover and actually prevented downtime. All in all, the downtime we have expereinced, IMO, has been much greater than what we'd have with 3 standalone PC based HMI systems.

My counterpart has spent countless man hours patching, tweaking, studying and developing the HMI stuff. He has worked with Rockwell and tried all sorts of arrangements related to which servers host which peices of the jigsaw puzzle, but no matter what, there have been problems.

We have decided to go to a standalone system, so that downtime will at least be limited to 1/3 of the plant, and will be more manageable for us controls folks without so much dependence on IT.

Mike (my counterpart and our SE guru) is looking into RSViewSE standalone so that he won't lose all of his development. He frequently says he wishes he'd pushed mgmt to use wonderware or Ifix at the beginning, but now it would be like starting over to do so.

What I am wondering is would it be better to use RSView32, and would it be possible to re-use any of the RSViewSE stuff if we did that?

Also, what is the difference between:
FactoryTalk View Site Edition Station (ie 9701-VWSB100AENE)
FactoryTalk View Machine Edition Station for Windows (ie 9701-VWMR075AENE)

Thanks,
Paul
 
Sorry to hear you are having so many problems.

I wonder if CPR9 (5.0) would help?

We put in a CPR9 system 6 months agos with 2 redundant sets of HMI/DATA servers with 3 clx processors and running 8 clients plus 2 engineering stations and haven't had any issues to speak of. The system has about 150 motors and 300 instrument loops/indications. My one gripe is that the FT Alarm and Events doesn't support redundancy yet, so I couldn't get away from tags totally since I needed HMI tags to operate my alarm summary.

I'm suprised when I read about someone having so many problems, since our installation when so good.

I wonder if the architecture has something to do with it? System architecture could probably make or break you.
 
i dont think factory talk view will help you as there is no server-side VBA support at present. FactoryTalk SE is the SCADA-type application.

I dont have much practical experience with RSView but am fresh from Rockwell training! ... I think i learnt VBA scripting in RSView is single threaded, I also know that means there is no point in using 'DoEvents' in your loops and that can cause issues in VBA...run that lot by your guru and good luck!

I take it if rockwell are involved you know about log files etc? All latest service packs/patches are on. Your best hope might be to enhance the diagnostics as far as possible with regards your redundancy and sit tight & wait.
 
Last edited:
Yes, Mike is well on his way to having the application ported over to standalone versions. Rockwell is supposed to swing a deal for us on licensing, so yes, they are aware of our problems, approved of the original design, etc.

As far as I know, his VBA is not the root cause of our problems. He is using VBA to create text files which are sent to an external application for processing, receive and then parse the results and write them into PLC registers, then verify that the writes actually went through. This takes place once every two to ten minutes throughout the day.

All the latest patches for 4.0 are installed. He doesn't feel that continuing to upgrade by going to FT 5.0 will solve the over all problems after having read all the release notes on the subject.

Mike has tested the application in a standalone mode and with some minimal debugging of some tags, he had it working for one shift. Now its just a matter of getting the PC hardware and licenses purchased.

Thanks for all the comments.
Paul
 
What I am wondering is would it be better to use RSView32, and would it be possible to re-use any of the RSViewSE stuff if we did that?

Also, what is the difference between:
FactoryTalk View Site Edition Station (ie 9701-VWSB100AENE)
FactoryTalk View Machine Edition Station for Windows (ie 9701-VWMR075AENE)


1. RSView32 - would require re-egineering.
2. FactoryTalk View Machine Edition Station for Windows - would require re-egineering as ME doesn't support macros, alarming is differnt and some button functions different etc etc.
3. FactoryTalk View Site Edition Station - is the go!!

I have a customer when I am using 2 x FactoryTalk View Site Edition Stations both setup exactly the same. One on the plant floor and the other in supervisors office. Simple redundancy!! Only downside is PLC gets more comms load and alarm ack/reset etc needs some thinking through. But it works for this site were no IT support is readily avaliable.
 
Quick Update

Well, we got our distributor involved, and he worked with Rockwell to get us some help.

Basically, they claim that SE Distrubuted 5.0 fixes much of what's wrong with 4.0, and that they would come to our plant and be able to figure out what we are doing wrong (free of charge).

So they came in for two days and went over our network. They came up with a hit-list of things we need to change and provided Mike with 5.0 and advice on the conversion process.

Our IT guys gave him another server to work with, and VMWare so that he can test the application in 5.0 with one server running multiple VMWare sessions to provide the architecture we wanted.

So, after Mike got through some hurdles to get the system to run (not in production) and was happy that everything was working, he now has one of the floor clients (in production) pointed at his virtual server.

So, we are going to stick with SE Distributed redundant for now, and give 5.0 a chance. At some point, when Mike is ready, we will update all the original servers to Windows Server 2003 and install 5.0 on them and move the whole system forward.

We still need to explore the settings of some of our network switches. One of the biggest issues the RA experts found was using auto-negotiate. They said it is always better to give all the devices a fixed speed and duplex, so we have done that for the SLCs and the PCs, and some of the switches.

I have not been into the nitty gritty details, and this is still a work in progress. When (if) it's all said and done, I will try to post back again.

Paul
 

Similar Topics

I have (or will be getting) a 250A motor that I am building a starter for. The starter will be in 36x36NEMA enclosure with a disconnect. This...
Replies
5
Views
1,640
I have problem with plc siemens that continue alarms "Emergency circuit not fed, how to fix it ? Thanks
Replies
4
Views
2,075
I started a new thread for this subject. I just tried this... I went to the Rockwell Automation site / Activation support /Get new activations...
Replies
6
Views
10,565
Is it just me or is anyone else completely fed up with all this nonsense??? I paid good money recently for RSLogix500, RSLogix5000 and...
Replies
176
Views
99,780
Hi All I Have a power quality meter that I have previously used on A DOL motor as a data logger. I want to do the same on two inverter fed...
Replies
4
Views
1,803
Back
Top Bottom