CompactLogix status switching in and out of Connected:Browse Pending

We have a system with 4 CompactLogixs v20 and they keep changing from Connected: Protocol: EIP - Run Mode to Connected:Browse Pending then back. They are using the Legacy Allen-Bradley CompactLogix type and I have automatic browse turned off. They do not all drop out at once and probably average one PLC dropping per minute. I set up a contentious ping and was probably getting one time out in 50. The rest were ~1ms. The only thing that has changed is the network card on the server. It was happening before but only a couple of times a day. In trying to fix it I updated from 7.9.3 and updated java as well. They did not seem to help(But I was planning to do it soon anyways).
This is about the first message that shows in the logs:
“ABControlLogixTransportPool 30Dec2017 12:19:57 [TruckA] Reseting CIP connection for transaction ID of 7584”
Then it will say this:
“CompactLogixDriver 30Dec2017 12:20:01 Comm state changed to INACTIVE” then the comm state goes active with the same timestamp.
The next is:
“ABControlLogixBrowseRequest 30Dec2017 12:20:01 [TruckA] Creating browse request using HIGH priority.”
After that it goes back to working just fine.

Any help would be great. I will continue to work on it and will give an update if I find a solution.

Java 1.8.0_151-b12
Windows server 2012 64bit
7.9.5 64bit 8gb ram
4-CompactLogix PLCs v20
11,422 tags
2 to 6 clients (2 at the moment)
2-MSSQL databases

Not really enough detail to be sure, but it sounds like you are at or above the communcation’s capacity of the processors, possibly intermittently due to other devices querying the same PLCs. Take a look at the overall capacity % on the devices’ status pages. Consider increasing the communications time slice in the processor properties, and/or converting your PLCs’ continuous tasks to scheduled tasks (highly recommended).
Note that PLCs in the 5370 family (like the L24) are dramatically faster than their predecessors (like the L23) and can handle many more tag subscriptions. Also beware of the very last v20 firmware, as it implemented the same comm security limits as v21 firmware, and you’d have to switch drivers.

The CompactLogix are L45s with ENBTs and at v20.11 because the 20.3 security issues.
I thought of the same thing with the ethernet cards possibly being maxed out but the ENBTs are at 25% to 50% CPU. And The overhead slice is at 30%.
I also agree that we should probably switch over to timed tasked but I do not feel comfortable doing it without being onsite.
Thank you for your reply.

I wouldn’t expect the ENBTs processing load to be the limiting factor. The main CPU’s load is. Please also look at Ignition’s status page for each driver.

I moved Ignition from the server to a windows 7 pc. And now all of the issues are gone. So it appears to be something in the server or hosts network. I think that is where we are going to leave it until Tuesday.
Thank you for your replies.

I had a similar issue on some L16 compacts running v24. In the end restarting opc cleared it up.

In the end I ditched the ignition built in diagnistic ( goos but not enough detailed informarion) and built a new datatype and used a gsv to pull more information. It was fairly simple, the data structure for the gsv is laid out pretty well in the rockwell help files.

1 Like