SCENARIO:
A server running Honeywell management software connecting to an old DEC terminal server which is connected to various serial devices (industrial systems such as fire alarm panels, card readers etc).
I need to replace a number of these units with Cisco 2511 routers configured as terminal servers. I have a basic configuration consisting of an IP address on the e0 interface, baud/stopbits/parity etc set on the lines I need to use, and "transport input all" on those lines also. I can connect various DCE and DTE serial devices to the router and telnet to port 2001+ successfully and communicate with those test devices.
PROBLEM:
However, when I take the old DEC termserver offline and insert the cisco device (after changing the IP of course) the management software fails to connect to the end devices.
TROUBLESHOOTING:
If I do a "sh line" I can see that lines 1 - 8 are being accessed by the management software, so the network side of things would seem to be okay (no ARP cache issues etc I wouldn't think).
Inserted a serial tester between the router and the end devices and I believe I have the right cabling. Only rs-232 pins 2,3 and 7 are in use (Tx, Rx, GnD). I tested another device that uses only 2, 3 and 7 and it worked fine with the cisco router. This leads me to believe the serial end is okay also. I tried the cisco DCE connecter and a gender-bender on the serial device just for shits and giggles and still no luck.
The serial devices have been designed such that they don't respond with data unless they are asked specifically for it which means I can't simply reverse telnet to port 2001 on the router and expect to get text back from the end devices which is a BUMMER. I know that the devices only respond to requests from the honeywell software because I performed an ethernet packet capture on the network side and analysed the data. The request/response structure varies each time and is possibly encrypted somehow. I haven't tried to manually reproduce it yet...
I also have not yet gotten an ethernet packet capture between the software and the cisco router on the network side because at any one time I only ever have a few minutes to "break" these criticial systems before I need to put it all back online again.
Getting desperate! What have I missed?
A server running Honeywell management software connecting to an old DEC terminal server which is connected to various serial devices (industrial systems such as fire alarm panels, card readers etc).
I need to replace a number of these units with Cisco 2511 routers configured as terminal servers. I have a basic configuration consisting of an IP address on the e0 interface, baud/stopbits/parity etc set on the lines I need to use, and "transport input all" on those lines also. I can connect various DCE and DTE serial devices to the router and telnet to port 2001+ successfully and communicate with those test devices.
PROBLEM:
However, when I take the old DEC termserver offline and insert the cisco device (after changing the IP of course) the management software fails to connect to the end devices.
TROUBLESHOOTING:
If I do a "sh line" I can see that lines 1 - 8 are being accessed by the management software, so the network side of things would seem to be okay (no ARP cache issues etc I wouldn't think).
Inserted a serial tester between the router and the end devices and I believe I have the right cabling. Only rs-232 pins 2,3 and 7 are in use (Tx, Rx, GnD). I tested another device that uses only 2, 3 and 7 and it worked fine with the cisco router. This leads me to believe the serial end is okay also. I tried the cisco DCE connecter and a gender-bender on the serial device just for shits and giggles and still no luck.
The serial devices have been designed such that they don't respond with data unless they are asked specifically for it which means I can't simply reverse telnet to port 2001 on the router and expect to get text back from the end devices which is a BUMMER. I know that the devices only respond to requests from the honeywell software because I performed an ethernet packet capture on the network side and analysed the data. The request/response structure varies each time and is possibly encrypted somehow. I haven't tried to manually reproduce it yet...
I also have not yet gotten an ethernet packet capture between the software and the cisco router on the network side because at any one time I only ever have a few minutes to "break" these criticial systems before I need to put it all back online again.
Getting desperate! What have I missed?
