"no response" or "incomplete message" on receiving

iynehz commented 4 months ago

Versions

Python: 3.12
OS: Windows
Pymodbus: 3.6.x including 3.6.4 which is latest
Modbus Hardware (if used):

Pymodbus Specific

Server: rtu - sync
Client: rtu - sync

Description

It's a random issue. It's similar as https://github.com/pymodbus-dev/pymodbus/issues/2018 except that I use RTU and my error message is like

ERROR Modbus Error: [Input/Output] No Response received from the remote slave/Unable to decode response

or

Modbus Error: [Input/Output] Modbus Error: [Invalid Message] Incomplete message received, expected at least 4 bytes (2 received)

Version 3.5.4 is good.

Code and Logs

2024-02-18 16:53:13,523 DEBUG logging:103 Running transaction 8
2024-02-18 16:53:13,523 DEBUG Running transaction 8
2024-02-18 16:53:13,523 DEBUG logging:103 SEND: 0x11 0x6 0x0 0x0 0x66 0x66 0x20 0xd0
2024-02-18 16:53:13,523 DEBUG SEND: 0x11 0x6 0x0 0x0 0x66 0x66 0x20 0xd0
2024-02-18 16:53:13,524 DEBUG logging:103 Resetting frame - Current Frame in buffer -
2024-02-18 16:53:13,524 DEBUG Resetting frame - Current Frame in buffer -
2024-02-18 16:53:13,525 DEBUG logging:103 Changing state to IDLE - Last Frame End - 1708246393.515018 Current Time stamp - 1708246393.524001  
2024-02-18 16:53:13,525 DEBUG Changing state to IDLE - Last Frame End - 1708246393.515018 Current Time stamp - 1708246393.524001
2024-02-18 16:53:13,525 DEBUG logging:103 New Transaction state "SENDING"
2024-02-18 16:53:13,525 DEBUG New Transaction state "SENDING"
2024-02-18 16:53:13,526 DEBUG logging:103 Changing transaction state from "SENDING" to "WAITING FOR REPLY"
2024-02-18 16:53:13,526 DEBUG Changing transaction state from "SENDING" to "WAITING FOR REPLY"
2024-02-18 16:53:13,816 DEBUG logging:103 Transaction failed. (Modbus Error: [Invalid Message] Incomplete message received, expected at least 4 bytes (3 received))
2024-02-18 16:53:13,816 DEBUG Transaction failed. (Modbus Error: [Invalid Message] Incomplete message received, expected at least 4 bytes (3 received))

janiversen commented 4 months ago

Your device do not respond, so not much we can do. You do not share your code, so I cannot tell you if any of your parameters look suspicious.

A question are you sure your device is slave 17 ? (asking because it is unusual).

janiversen commented 4 months ago

Seems solved.

iynehz commented 4 months ago

Sorry was busy on something else and did not notice your earlier response. No it's not solved. I have multiple modbus devices on the bus that's why its address is not 1. The device is a MCU running Rust https://github.com/alttch/rmodbus library. As I said it's good under version pymodbus 3.5, but the randon failure is obvious under 3.6. So it's probably not something about my setup.

janiversen commented 4 months ago

I really cannot help you a lot...you do not share the code, so it's impossible to see if you changed the parameters. You are using the sync version, which she had very few changes.

Also 3.6.4 is a lot faster which might influence your device.

janiversen commented 4 months ago

This comment "So it's probably not something about my setup." is also valid for the pymodbus library 😄 , the library just does what you ask it to do.

What you can do is to try and use the async version, which is a different code part, so it might give another result.

And just for the record this issue have very little in common with #2018, which is async, tcp and the problem is receiving multiple packets.

janiversen commented 4 months ago

Closing due to lack of information.

pymodbus-dev / pymodbus