I've been testing 6.1.1 controller redundancy with ModBus_TCP I/O and loving it. Terrific! My question is about automatically switching displays to the ACTIVE controller on dedicated Operator Consoles. After thinking through possible schemes to make graphical switching automatic, I'm soliciting recommendations and design guidance.

/Ron

Hi Ron,

The idea is that you should create a separate operator station. This will connect to both primary and secondary node and automatically switch to the currently active node.

/Claes

Claes,

Yes, thank you and Happy New Year. I have built redundant controllers and currently testing controller fail-over from a separate operator station perspective. Still working to understand all of the nuances before putting into production. I am using $MountObject as per usual and did remember to include Hierarchies in the RedcomPacket. I see there is another $MountDynObject, but don't know what that's about and find no detailed documentation about it?

One minor question, please. From the Workstation, distributing to a primary and secondary controller works fine. But when distributing from the Workstation to an operator station, a distribution error occurs. The .tgz file downloads to the operator station properly, but fails to unpack it. The error message shows bash: line 1: /pwr_pkg.sh: No such file or directory. Yet, from the Operator station machine's home directory ~/, I can run> pwr_pkg.sh -i project.tgz and the script works perfectly.


Doesn't the Workstation Distributor ssh to the target node and run pwr_pkg.sh locally on the target node, where pwr_pkg.sh is embedded within the .tgz? On the Operator Station, the script is also found in $pwr_exe/pwr_pkg.sh with owner pwrp:pwrp and execution privileges. In this case, it appears the distributor is confused on the directory name when trying to run pwr_pkg.sh? I'm not sure where to look for the problem, but suspect it's a minor configuration issue? Do you suspect on the Workstation in the project directory or the Operator Station project volume, or the Operator Station itself or somewhere else?

Your insights are much appreciated.


Regards,
Ron

Hi Ron,

$MountDynObject mounts a dynamic object in a dynamic volume. It can be used in storage stations to mount objects on the system volume under pwrNode.

When logging in to a operator or process station, the environment is normally set up by /etc/pwrp_profile. I've seen on rpios for example that when executing a single command via ssh, this file is not executed and then $pwr_exe is not defined and pwr_pkg.sh will not be found. I think it can be fixed by adding the line

at the beginning of /home/pwrp/.bashrc.

/Claes

I had added source /etc/pwrp_profile at the bottom of /~.bashrc. When moved to the top, distribute works perfectly.

Thanks!
/Ron