OPC Communication Issues

Please be gentle as I am rather new to this..

I am doing data collection from 6 PLCs (5 SLC5/03,1 SLC5/04) via OPC. The OPC server is RSLinx and the data is being collected with an update interval of 1 second. The PLCs are on ethernet via an AB NetENI device. I am collecting approximately 30 data points per PLC (most data points are a word).

The configuration works well 90% of the time, providing 1 second updates of the data, But occasionally (lets say anywhere from 10 seconds to 2 minutes, every 1-2 hours) we get bunch of bad data (the OPC data Quality indicator is "Bad" and the values are all 0). Upon looking at the communication events logged by RSLinx, I found the following two different errors for many different data points....

MYSAMPLENET\111.11.11.11 : N99:0 : 08/10/06 - 13:25:28 : 05h
Items: N99:0, N99:1, N99:10, N99:11, N99:12, N99:14, N99:2, N99:3, N99:4, N99:5, N99:6, N99:7, N99:8, N99:9
Type: READ
Mfg: Allen-Bradley
PlcType: SLC-503+
Desc: Local Error: TimeOut
The application layer time out while waiting for a response.
Number of occurrences: 670
Last Error occurred: 08/16/06 - 10:41:54


MYSAMPLENET\111.11.11.12 : N99:0 : 08/10/06 - 15:15:11 : 02h
Items: N99:0, N99:1, N99:10, N99:11, N99:12, N99:14, N99:2, N99:3, N99:4, N99:5, N99:6, N99:7, N99:8, N99:9
Type: READ
Mfg: Allen-Bradley
PlcType: SLC-503+
Desc: Local Error: No Ack Received!
The remote station specified does not acknowledge the message. Check if the destination station is on the Data Highway/PCL link.
Number of occurrences: 504
Last Error occurred: 08/16/06 - 10:38:07


As an aside, we tried testing with Kepware OPC server and didn't get any Bad data messages from the OPC server, but we would only get updates every 4 seconds (despite the 1 second update rate) if we collected from all 6 PLCs.


Does anyone have any ideas as to what might be causing such behaviour?

Thanks very much.

AAARGH!!!! OPC!!!!! Not impressed from my experiences with it. Bad data seems to be the flavour of the day with the 3 OPC servers I have been forced to use. Maybe it will improve. It could also be the implementation that is causing the problem.

Good luck!!!! I try not to use OPC if I can help it.

I am sure someone will have some happy experiences to relate.

No experience with NET-ENI's but I have heard that you should not expect exceptional performance from these.
As a test, what happens if you extend the update time to 2 secs ?
If you dont get the bad data's then it does look like you are logging too much or too often.
I use OPC a lot, but mostly for event driven data transfers, not cyclical logging.
If you have a performant connection (direct ethernet for example, not via NET-ENI) then I have no problem with cyclical logging at relatively high rates.