My setup here in the plant has separate Ignition OPC servers at each line, and our “main” gateway attaches to them. Hub and Spoke, but OPC-only.
If one of the spokes go down, like on a power cycle, it won’t reconnect properly when the remote comes back online. Picking an endpoint, or disabling/re-enabling the connection puts things back in order, but it’s kind of a pain in the keester.
Running 7.8.1 on the Hub, 7.7.0 or .1 or .2 on the spokes.
This worked okay pre-7.8. Do I need to upgrade the spokes as well?
You can send your logs in so we can double-check, but there was one last issue that looked like this fixed in 7.8.2. If you’d like an early copy of the module let me know.
Thanks for the info, Kevin.
If you can float me a copy of the module to try out, that would be great.
Although it would be even greater if our maintenance guys would power cycle a line as a last resort and not a first option… 
Just a quick revisit.
7.8.2 didn’t do anything for me, but installing 7.8.3 on a new VM (since the apt-repositories aren’t available anymore, it was most painless way to make sure there weren’t any other linux-y hiccups) seems to have done the trick.
Not sure if it was the version or the fresh install that did it for me, but I’ll be quiet about it for now. 
There ended up being additional bugs fixed in 7.8.3 related to this. The only known/outstanding issue now is with connections to Kepware, but it’s a bug on their side that we’re waiting on.
Huh. I guess I spoke too soon… 
While it appeared that things recovered okay, the actual OPC values never update.
Here’s the log. Around 5:00 pm or so is where I cut power to the spoke. I don’t quite remember when I plugged the spoke back in. Around 8:30pm is where I noticed things weren’t updating and reset the connection.
The connection is named 31CX. Its IP address is 192.168.140.64
And the saga continues… I knew I should have made some theme music for this…
Today, the OPC connections seemed to be working okay, but expression and query tags had crapped out. log is attached, but I have no data as to when this happened.
Restarting the default tag provider got thing back on track, but this is starting to kill me!
Also, running 7.8.4 if it helps. Running that even if it doesn’t. 
Thanks, guys!
I think this may be most easily solved by upgrading the “spoke” systems you’re connected to. Just to the latest version of 7.7, you don’t need to go to 7.8.
If it continues after that, on the central/client machine turn loggers ClientChannelManager and ClientSessionManager to DEBUG and then get the logs after something happens again.
Kevin,
I seem to be having a similar issue with a 7.8.2 setup. The OPC connection from server to server is over a cell modem connection and every two months the signal gets to low, and the connection fails. Even after the cell connection comes back, the OPC connection does not until I restart it.
Should I update both sides to 7.8.4?
[quote=“Caleb Bohon”]Kevin,
I seem to be having a similar issue with a 7.8.2 setup. The OPC connection from server to server is over a cell modem connection and every two months the signal gets to low, and the connection fails. Even after the cell connection comes back, the OPC connection does not until I restart it.
Should I update both sides to 7.8.4?[/quote]
You should at least update the client side, but doing both sides is always better. If it continues to not work the same loggers I mentioned previously would need to be enabled.
I have upgraded to 7.8.4 will repost if I continue to have problems.
It looks like my upgrade to 7.8.4 fixed the problem with the OPC connection. Now if only I could keep the cell connection from dropping…