Problem with subscribe_data_change

[JuliaCordeiro]

We are using the subscription.subscribe_data_change method passing these two nodes:

• ns=3;s=167.252.15.79->OPCServer.WinCC.1>CPY_R0_PRJ_OSS1SB::Eq_Com_CFL_L1/AIs_1_CFL_L1/TT36_10_01/MEAS_MON.U
• ns=3;s=167.252.15.79->OPCServer.WinCC.1>CPY_R0_PRJ_OSS1SB::Eq_Com_CFL_L1/AIs_1_CFL_L1/CT36_10_01/MEAS_MON.U

These two nodes have their values changed every second on the OPC Server, but we are only receiving datachange_notification for the second node with the value of the first node.
Example:

Tag_1 = {
address: ns=3;s=167.252.15.79->OPCServer.WinCC.1>CPY_R0_PRJ_OSS1SB::Eq_Com_CFL_L1/AIs_1_CFL_L1/TT36_10_01/MEAS_MON.U,
value: 30,8
}
Tag_2 = {
address: ns=3;s=167.252.15.79->OPCServer.WinCC.1>CPY_R0_PRJ_OSS1SB::Eq_Com_CFL_L1/AIs_1_CFL_L1/CT36_10_01/MEAS_MON.U,
value: 0,11
}

Output at subscription: Tag_2.value = 30,8

[AndreasHeine]

can you provide more logs?

what are the subscription parameter e.g. publishInterval?

what params do you pass to subscribe datachange? e.g. samplinginterval, filters, ...

small code snipits would help too!

have you tested it with UaExpert does it work with it?

[JuliaCordeiro]

We are creating the subscription inside the subscribe function of OPCUA_SUB_HANDLER client.
This is example with some pieces of our code:

self.polling_interval = 128
self.client = Client(opc_ua_host_address)

subscription = await self.client.create_subscription( period=self.polling_interval, handler=self )
await subscription.subscribe_data_change(self.nodes)

And about the OPC, we're using OPCExpert with OPC_UA protocol

[AndreasHeine]

have you tried a simplified test like in our examples?

e.g.:
https://github.com/FreeOpcUa/opcua-asyncio/blob/master/examples%2Fclient-subscription.py
https://github.com/AndreasHeine/opcua-tutorial/blob/main/client%2Fsubscribe.py

[JuliaCordeiro]

Yes, we're just logging inside the datachange_notification and doing a reset of our watchdog.

def datachange_notification(self, node: Node, val, data):
self.logger.log_info( f"DataChange Node: {node.str()}, val tag_reading: {val}", 200)
self.watchdog.reset()

[JuliaCordeiro]

We've noticed that in a very strange way, the test_tag that we insert at the first position of our nodes_array was causing the problem. But we couldn't find the reason for this behavior.
Example of our node_array with problem:

nodes = [
"ns=0;i=2258",
"ns=3;s=167.252.15.79->OPCServer.WinCC.1>CPY_R0_PRJ_OSS1SB::Eq_Com_CFL_L1/AIs_1_CFL_L1/TT36_10_01/MEAS_MON.U",
"ns=3;s=167.252.15.79->OPCServer.WinCC.1>CPY_R0_PRJ_OSS1SB::Eq_Com_CFL_L1/AIs_1_CFL_L1/CT36_10_01/MEAS_MON.U"
]

Change the position of the test_tag for the middle of the array the error persist.
We find out that using the test_tag at the last position the problem, apparently, is resolved. We would like to know why.

[AndreasHeine]

the thing is we just ask the server to add nodes (monitored items) to a subscription with a few parameters... we as client have no clue what the win cc opc ua server does with that!