How to remotely reboot IXXAT gateway

I have an Ixxat SG gateway that timed out when trying to save changes to the web PLC can that now I cannot log into. The strange thing is the device is online (I can ping it, I can read values from the OPC server, etc) but when I go to the web UI it loads the initial home.html page then displays an error “Timeout exceeded” after a while.

This device is 3000 miles away and I cannot remotely power cycle it, what options do I have to get back into the device remotely?

There isn’t really an easy way to do this…

Have you tried clearing the browser cache? a different browser or entirely different PC?

Can you see the device in CHIPTool?

I didn’t think of CHIPTool or trying a different PC. I really need to get into the device so I had someone local powercycle the panel and now the whole thing is down and won’t come back up. This is going to be expensive to recover from.

For future reference, is there some way to enable ssh access to the device to reboot it?

The only thing it supports for remote connections is ftp, telnet, http(s). I will have to look into this more to find out how you would initiate a reboot for this device.

Adam,

Do you know which firmware is on the gateway? Have you been able to reconnect?

Kyle

Hi Kyle

This setup hasn’t gone into production, yet, so I had someone local disconnect and overnight all of the hardware to me. I will have it here later today and and investigate further at that time.

Adam

OK good. Please check the firmware version and let me know when it comes in.

Kyle

I got the hardware a little while ago and hooked it up here. The controller which had locked up and started this whole issue has the following characteristics:

image.png

It appears to have the latest firmware. Can you share the logs?

Sure, but the log just goes back to when the device was last restarted which was this morning.

eventLog-2019-08-29-08-16-45.csv (3.35 KB)

Hi, thanks. I figured there wouldn’t be much, but thought I’d take a look anyway. Does the device appear to be running normally again? What changes were you making when the web server locked up, and did you happen to notice if they were stored or not?

I am still waiting to hear back from developers about the remote restart question.

Hi Kyle

The changes were adding some logic to the WebPLC program and were saved, the device is running normally again (the thing that took the site down when then panel was powercycled was a problem with the router).

Now the device has another issue, which is that one of the Modbus slaves is not readable:

I’ve set up a debugger on the RTU slave (a Moxa MGate MB3170) and see requests going to the other slaves (represented on the RTU link as different slave IDs) but not that one (slave ID 1). The weird thing is this was working earlier today! I changed the serial link speed then set it back, rebooting in between, otherwise nothing has changed. Is this some sort of port locking problem on the serial port?

Is this RS485?

How is it wired? Keep in mind if a device in RS485 mode is last in line in the bus topology a termination is required.The termination resistor needs to be 120 Ohm and must be inserted between pin 3 and pin 7 of the serial interface.

It is 485, but in this case the termination resistor was not the issue - it was another HMS device. Specifically, the IXXAT SG connects via 2 wire RS485 to a Modbus serial to TCP gateway which provides network isolation into the internal power plant SCADA network. This Modbus RTU to TCP gateway in turn talks to several Modbus TCP slaves re-writing the requests coming from the IXXAT (RTU master) as TCP requests and re-writes the TCP responses from the slaves as RTU responses which the IXXAT gets as if the slaves were RTU and connected to the IXXAT on the RS485 bus. The Modbus RTU to TCP gateway maps different slave IDs on the serial side to the TCP/IP slave devices on the ethernet side.

Because we re-located the panel back to my lab I had to configure some hardware to stand in as the slaves the system normally talks to in the plant. I used two PLCs for this and an older HMS AnyBus SG device; after running tyhe RTU to TCP gateway in debug mode I saw that the HMS slave was not responding to the requests from the master fast enough (or at all in some cases) so I replaced it with a PC running a Modbus slave application to stand in as the same plant device and now everything works.

This was particularly hard to troubleshoot because there is no way (that I am aware of) to see a log of the actual bytes sent and received through the serial port on the IXXAT SG device. That would have allowed me to see that the requests to that particular slave ID (mapped by the Modbus gateway to a given TCP slave) was not responding.

I’m sorry the diagnostic options for the SG Gateway are limited. I have also confirmed that doing a remote reboot is not an option unfortunately. Please let me know if you have any other questions.