We have some problems with Ignition OPC UA server. First some information about used software and versions:
Windows Server 2008 R2
Ignition version: 7.9.1
Java version: Oracle Corporation 1.8.0_40
When we are reading large chunk of data (1063 boolean values) we are receiving BadTimeout exception. Reconnection is done successfully by the client 5s later. Client is reading data every 500ms. Bellow is one screenshot with the exception. Timestamp of exception is the same as in our client service log.
We are connecting to Allen-Bradley 1769-L36ERM with firmware 21.11 with Allen-Bradley Logix Driver. Timeout happens couple times a day, but without clear pattern.
We had this issue on a project and it was like pturmel suggested. I had tried to set the scan rate for the tags being read to be too fast so not all values were returned before the next read was supposed to happen.
To solve that, I put the high priority tags into their own scan class that was at the faster rate and then split the others among a couple of slower scan classes.
Long and short. You might benefit by splitting up the booleans you are reading among a couple of scan classes rather than just the default scan class.
First attempt, try changing the rate to 750ms instead of 500ms and see if the incidents decrease in frequency.
Oh, and that 5s reconnect is defined on the connection settings page if you care.
@Kevin.Herron - yes default settings are used. What value could improve the situation? What is timeline for 7.9.11 release?
@BillS - we are using OPC UA server with our custom developed .NET windows service (we are not using Ignition Designer, etc…). We tried changing time rate from 500ms to 5000ms but to no avail. We also reduced number of tags that are being read from 1000 to 200 but the same error occurred - only less frequent.
7.9.11 should be available in a couple weeks? The beta is available now.
I’m not really sure what’s going on here. A TimeoutException from SynchronousRead or SynchronousWrite, as you’re seeing, generally indicates a bug in the driver unless that driver’s communication timeout had been configured to be longer than 60s.
I’ve never seen this happen with the newer Logix driver so I’m just hoping some bug fix that occurrred in the subsequent 10 releases fixes it.
Aside from a bug, the alternative is that the device is occasionally so loaded it takes longer than 60s before the read request even gets executed. Does your client wait for the result of a read before issuing the next? What’s the load factor on the diagnostics page for this device? (I think 7.9.1 has this page… not sure)
