| T.R | Title | User | Personal
Name | Date | Lines |
|---|
| 2462.1 | | NETCAD::GERHARDT | Out of Control | Mon Jul 03 1995 14:36 | 14 |
|
!1. How did it get to that state, and how do I prevent myself from
!getting there again?
It could be that the power was shut off to the MAM before the
code was written to the FLASH memory or that there is bad FLASH.
If the boot up code finds that the checksum on the image stored in
FLASH is different from what it should be it goes into that mode.
!2. Can I reload the firmware into this MAM, or do I need to send it
!back and get another one?
you can reload it by connecting the OBM port up to a SLIP line and
doing a TFTP put to the port.
|
| 2462.2 | | STRWRS::KOCH_P | It never hurts to ask... | Mon Jul 03 1995 20:21 | 3 |
|
did you make sure that the TFTP/Bootp server was up and running? Did it
show that it loaded the MAM?
|
| 2462.3 | | GRANPA::DFAUST | Bad Things, man... | Tue Jul 04 1995 17:24 | 6 |
| re: -1
Yes... I had just upgraded two other DEChub 900s prior to this one.
I'll try the advise from .1 tommorrow, when I return to work.
Dennis Faust
|
| 2462.4 | What's causing BOOTP | NETCAD::FORINO | | Wed Jul 05 1995 11:56 | 5 |
| Note that the HUBloader only relies on the TFTP Server to load modules.
BOOTP is not needed from its point of view. The question is what is
getting the MAM in the state issue a BOOTP request.
John
|
| 2462.5 | Same problem;reporting it to MAM team | NETCAD::ROUNDS | | Wed Jul 05 1995 17:32 | 9 |
| Dennis,
I had the same thing happen to me with 2 MAMs. All that day I was
having problems with TFPT timeouts.
I work in the hub engineering group so will let you know what I find
out.
Nat
|
| 2462.6 | Could we DOCUMENT this in customer docs? | PTOJJD::DANZAK | Pittsburgher � | Thu Jul 06 1995 09:43 | 21 |
| Folks - could we DOCUMENT this -in the CUSTOMER documentation! If you
don't do it the following sequence happens:
- Customer is out of service
- Customer logs service call (DEC $ start getting spent)
- Digital field engineer orders swap-out of hub (more digital $ get
spent)
- After a bit, customer is re-running
- We have spent lots of Digital $, expended $ etc. that the customer
could have fixed if we outlined the procedure!
I was at a cutstomer site in Eastbejeesus where we zapped a repeater
(i.e. had a hardware load problem etc.) and ended up toasting it.
Because nobody knew how to make it well (it was not documented) we just
did a return-to-Digital warrantly on it (costing Digital $ but not as
much as me driving 3 hours to get there again and fix it).
So can we DOCUMENT how to recover toasted modules? It IS, after all,
the thunderstorm season in many places...>!
j
|