Powerflex Net I/O Timeout

I am working with a 4 year old system with a Powerflex 755 that has suddenly developed a NET I/O TIMEOUT every hour on the hour (+/- 30 seconds). As far as I know, this means that something on the network is trying to communicate with the drive but the comm has failed long enough to timeout. The error log has "Net IO Timeout" immediately followed by "Network Open" every hour like clockwork. Since the event is set to FAULT, this throws a fault to the PLC and requires an operator reset. I tried just changing the parameter to STOP instead of fault but this still caused the system to get out of sequence but now without letting the operator know that we had the fault. The PLC is a L43B using Ethernet/IP to run the drive. There is a managed switch in the cabinet. Any ideas?

I am in the same boat as you but my installation is only a few months old. One thing I did find that helps is to reset/reboot the network switch. When I did that the faults stopped for about 24 hours. They came back but by then the customer was done using the system.

What switch are you using? I am using the Stratix 5700.

I am heading to the site Monday and will post back if I make any dicoverys.

A network fault that happens every hour on the hour sounds like a switch management setting.

Do you know if you are using Unicast or Multicast I/O connections ? Multicast group definitions are the sort of thing that might expire after an hour.

What model of managed switch is it ? Do you know if you've configured IGMP Snooping and Querying on it ?

Are there any other managed switches connected to it that might be forcing an IGMP Querier Election ?

Bullzi said:

I am in the same boat as you but my installation is only a few months old. One thing I did find that helps is to reset/reboot the network switch. When I did that the faults stopped for about 24 hours. They came back but by then the customer was done using the system.

What switch are you using? I am using the Stratix 5700.

I am heading to the site Monday and will post back if I make any dicoverys.

Click to expand...

Ken Roach said:

A network fault that happens every hour on the hour sounds like a switch management setting.

Do you know if you are using Unicast or Multicast I/O connections ? Multicast group definitions are the sort of thing that might expire after an hour.

What model of managed switch is it ? Do you know if you've configured IGMP Snooping and Querying on it ?

Are there any other managed switches connected to it that might be forcing an IGMP Querier Election ?

Click to expand...

Thank you for the reply. The switch is a D-LINK DGS-1224T. Unfortunately my customer does not have the password to access the switch at this time and it is not the default "admin"... They are trying to retrieve the password so I can look at IGMP snooping. So far I do see anything that is multicast but I will confirm that. I found out today that the problem possibly started with a server upgrade so we are looking at other switches as well. Thanks, I'll keep you updated.

Thanks again... It could be that this is outside of the system. I will ask their IT guy to check other switches etc.. We are not using multicast and the switch does not have IGMP snooping enabled. I am just walking into this process line. However, I was checking the processor for minor errors and there were a few program errors - code 4 arithmetic overflow (I did not write this program). The only instructions located where the errors referenced were DIVs. Since the errors all occurred at night when the line is not running, I cleared them. I am guessing that we are trying to divide by zero at night. After the errors were cleared the maintenance guy said the net I/O timeout problem went away for the rest of the day yesterday. They came back today and the errors came back again last night. I cannot imaging that the two things are related but this confuses me.

hi fellas. i would not start a new line.
im having the same issue "4037 net IO timeout"
o got here :
https://rockwellautomation.custhelp.com/app/answers/answer_view/a_id/475410

anyone has the same issue before ?

Welcome to the PLCTalk forum community !

Even though the basics of your question are similar to this thread, I am going ask you to provide the details and context for your specific installation and diagnostics on a new thread.

I have created it for you, and you should post updates there:



https://plctalk.net/qanda/showthread.php?t=135869