It’s no longer asking DHCP for an IP address. Arp-ping didn’t work either.
LED status with LAN cable plugged in, no serial cable:
Power: green
Lan TX: green
Diag: red
Serial CTS: green
The Lan RX led blinks green when I try and ping the board, but I don’t get a response.
If I disconnect the LAN cable, it’s showing:
Power: green
Link: red
Diag: red
Serial CTS: green
Any troubleshooting thoughts?
I’d first try sniffing with Ethreal to see if the Digi is issuing the DHCP request but is being rejected for some reason - or even annoucing a fixed IP in a gratuitous ARP. You’ll also see if there is a duplicate IP problem - Digi products don’t go off-line when assigned duplicate IP but if 2 products “fight” over an IP, then ARP can work very unpredictabley.
Otherwise, if you have a trained monkey handy you can try to do a factory reset. (Moneky only required for “new” hardware with that tiny reset button - hmm). In one of the vents just under the power LED is the button - older units have a nice large black button; newer units have a white button like 1 x 2mm. Hold this in while you power up the unit and keep holding untill the RED diag LED blinks out a pattern (30 seconds?). Personally, I don’t have a trained monkey handy so I remove my shoe and use my big toe to toggle the on/off switch of a power strip (sad but true :-|) That lets me hold the IA/IAP with one hand as well as accurately press the rounded reset button with my other hand and a plastic stick.
The button is to be replaced with a larger one next board spin early in 2006.
> I’d first try sniffing with Ethreal to see if the
> Digi is issuing the DHCP request but is being
> rejected for some reason - or even annoucing a fixed
> IP in a gratuitous ARP. You’ll also see if there is a
> duplicate IP problem - Digi products don’t go
> off-line when assigned duplicate IP but if 2 products
> “fight” over an IP, then ARP can work very
> unpredictabley.
Nothing. No ARP or broadcast traffic from the Digi at all.
> In one of the vents just under the power LED is the
> e button - older units have a nice large black
No monkey required–I’ve got the board out of the box, and it’s pretty easy to hold in. The first time I power cycled with the switch in, the diag LED did blink after 15-30s. No help.
I’ve attempted the factory reset three more times, but no diag LED blink…is the factory reset a one time only event? Wait, it just blinked. Next few times–no blink. Hmm. Every once in a while, the factory reset switch seems to blink the diag LED, but not every time. Very odd.
I did a quick scan through the pdfs, but I couldn’t locate the docs for the LED codes, or this factory reset. Can you point me to the right file?
What about this little 4 pin connector next to the switch? Is it some kind of serial port? Does that help me?
> The button is to be replaced with a larger one next
> board spin early in 2006.
And with, dare I hope, compliant PoE? 
(Sorry, I guess that horse is dead already.)
“Factory Default” by this button just wipes the NVRAM section of the config - forcing a fresh startup and all defaults in config. It can be redone as often as you like - however, there will be a “busy period” of up to 6 minutes while the unit recreates it’s SSL/TLS 1024-bit DH private & public keys. During this period the CPU is rather busy and HTTP support is unpredictable. I just do it and go get a cup of coffee. The exact time is a bit random because it depends on how “lucky” the algorithm gets in a first guess, since it then needs to hunt for a nearby 1024-bit prime-number.
Sounds to me like a hardware fault - check into our RMA procedure. Normally, on-constantly or a nice even 50/50 blink relate to IP problems (ie: no DHCP server, dup IP, etc)
During factory reset, you’ll see what we call a 1-5-1 pattern, with a short off-period in between. This repeats as long as the reset button is pressed.
Other LED blink codes ALL relate to low-level hardware issues that cannot be solved in the field. That 4-pin header is used during manufacturing to write-enable read-only sections of the flash (& other things) so not of any use in the field. I don’t even have the gizmo to use it since I don’t work in the boot-strap loader code.
And yes the PoE will be fixed in the next hardware to support both “spans”, however the plan is to move it to a special partnumber since the cost of PoE components is too high considering that 99+% of users don’t even consider its use. Just like how the TS4 (w/PoE) has been split into the TS4 and TS4P (with and without PoE respectively), we’ll need to create a new IAP (IAPP?) with the PoE populated.
> “Factory Default” by this button just wipes the NVRAM
> section of the config - forcing a fresh startup and
> all defaults in config. It can be redone as often as
> you like - however, there will be a “busy period” of
> up to 6 minutes while the unit recreates it’s SSL/TLS
I’ve let it sit over an hour, and the diag light never goes out.
> Sounds to me like a hardware fault - check into our
> RMA procedure. Normally, on-constantly or a nice even
> 50/50 blink relate to IP problems (ie: no DHCP
> server, dup IP, etc)
RMA isn’t an option. We had already removed half the connectors and were preparing to pot the unit when it was dropped. Strangely enough, there’s no apparent physical damage. Maybe ESD?
> During factory reset, you’ll see what we call a 1-5-1
> pattern, with a short off-period in between. This
> repeats as long as the reset button is pressed.
Haven’t been able to get the blink code to appear again.
Too bad I don’t have a coldfire BDM cable.