Updating pictory from 1.4.2 to 1.4.7 mess up

[DRi]

Hello,
I updated pictory from 1.4.2 to 1.4.7 and it looks like several things don't work as expected. (strecth image of course)
First of all we lost a part of the modbus configuration :

Capture d’écran du 2021-08-05 12-00-15.png (18.96 KiB) Viewed 6605 times

corresponding to

Code: Select all

,"mem": 
{"0": ["device_path","/dev/ttyRS485","256","410",false,"0312","INFO: value modified to adapt base device type",""],
"1": ["baud_rate","19200","32","442",false,"0313","",""],
"2": ["parity","0","8","446",false,"0314","",""],
"3": ["data_bits","8","8","447",false,"0315","",""],
"4": ["stop_bits","1","8","448",false,"0316","",""]
},

in _config.rsc,
where 0 to 4 content was replaced by things like that

Code: Select all

"Action_Status_Reset_18","0","8","409",false,"0311", "",""

I had to retrieve it from and old _config.rsc file
Did it happen before ?

My second concern is about the modbusRTU master configuration.
I wanted to disable the 32 rows limit.
It seems that even if the configuration is correct (for the slave) for every row after n°32 I don't retrieve any value (0) with piTest -r.

Capture d’écran du 2021-08-05 14-18-55.png (16.33 KiB) Viewed 6605 times

but before row 32 the exact same sensor gives the value as expected:

Capture d’écran du 2021-08-05 14-19-19.png (24.54 KiB) Viewed 6605 times

How can it be solved ?
Thanks for your help
DRi

[dirk]

Hello DRi, I am dismayed by the problems you report with the update.
There may be a known issue. Please have a look at this post:
https://revolution.kunbus.de/forum/view ... =10#p10760

[RR4711]

I think this is more related to the change in the RSC file when the the was some inconsistencies between process image and offsets regarding i think it was ACTION_RESET or so. I don't remember the details but there was stuff changed and potentially old config files no longer compatible.

[RR4711]

I mean this one: https://revolution.kunbus.de/forum/view ... 7282#p7320

[DRi]

Hello,
@dirk the solution you provided didn't make it this time.
So I tested again today with a fresh new revpi (on buster this time) first start.
I went to pictory (2.0.3) added a modbus master virtual 150 device.
Configured 32 lines in extended data for this device (mostly placeholders) everything worked.
Then I added a 33rd row (checking the disable row limit box) and when I tried piTest -r Input_Word_33 for the corresponding sensor I didn't retrieve any value.
But the exact same sensor which worked previously didn't send anything when I placed it before 33rd position in extended data (didn't appear twice in the list for each test).
I tried dirk's fix for pimodbus-master again, without better result.
As I'm not a german speaker @RR4711 , I tried to check the .rap file but didn't understand what to look for or how you find the kind of overlap neither.

DRI_issue_pictory.zip

(15.1 KiB) Downloaded 468 times

You can find details (rsc + rap) in the attachment archive, where in rsc files there are just placeholders + the last line is an actual sensor working fine.

As a last word here my use case and why I need more than 32 lines in extended data:
My revpi is installed in a remote location (send data via internet over gsm m2m) and monitors sensors + manages some logical functions.
Some sensors are probes which give ground Temperature + moisture for instance, so I need to read several register of the same modbus slave, I read 9 registers but just at one single slave adress. + To refresh the values I have to write in some registers of the same slave again.
This is how I reached 32 lines in the extended data part of the modbus master virtual device.
I'd like to find a solution before we send bigger sites in production, because I can't afford to drive several hours to manualy fix it afterwards.
Thanks DRi

[dirk]

Hi DRi thanks for the constructive and competent description and also for the community help, this is really great.
I learned from Amar that the only solution at the moment is to recreate the PiCtory configuration.
Pictory offsets for Some modbus devices were changed in RAP files between version 1.4.2 to 1.4.7

[DRi]

I can't make it work (not even on buster + pictory 2.0.3) do you have a corrected version of the .rap file ?

[dirk]

Hello DRi, here is the configuration I created with PiCtory V2.0.3.
Please make a backup of your system before you perform the following steps. For this you can have a look at the tutorial "Image":
https://revolutionpi.com/tutorials/vide ... =en_US#VT1

Because of the missing permission of user "pi" on the target PiCtory directory you can bring it to your system like this
1. copy this file to "/tmp" with this instruction https://kunbus-gmbh.atlassian.net/servi ... 1875444147
2. execute this command:

Code: Select all

sudo cp "/tmp/Configuration 33 Lines PiCtory V2.0.3.rsc" /var/www/revpi/pictory/projects/_config.rsc 

I hope that you can now successfully get your project up and running with this configuration.

[DRi]

Hello dirk,
Thanks for posting the config.
I tried again with a fresh install of my revpi connect and the buster image.
Loaded your config, made some minor changes to match what's currently on the test bench and I still can't make it work after line 32.
Even though in the rsc file everything seems ok regarding the last lines eg : ActionId_33_ModbusRTUMaster_20180406_1_2 and so on.
Did you make it work with the config you send me on your side ?
Regards
DRi