In article <. com>,
"sqrfolkdnc" <> wrote:
> I have three AGS+ routers in my home museum (at least the cover plates
> say AGS+). Two of them have a PC board above the metal plate at the
> top that is connected only by a ribbon cable and does not plug into the
> bus. The one (#2) without the board is a "minimal" system with only a
> CSC/4 and CSC/ENVM board and a CSC-2E2T board. The others have 5 io
> boards: #3 has FDDI, 6 ethernet and 8 (4+4) token ring, the #1 machine
> is 14 (2+6+6) ethernet and 4 serial. Both #1 & #3 have the CCTL2 board
> to support newer io boards.
The small board is an NVRAM or Flash memory board. I know the form
factor is right for the CSC-MC NVRAM board and wrong for the CSC-MT
which is a full MCI bus card, but I don't know about the CSC-M or the
CSC-MC+ (see
<
http://www.cisco.com/en/US/products/...eol_notice0900
aecd8029d3ed.html>)
It would be odd for it to be a CSC-MC because the ENVM board holds the
NVRAM for AGS+ systems.
> Questions/problems:
>
> 1) units #1 and #3 fail while booting, repeatedly, but after ~60
> minutes (#3) or ~75 minutes (#1), they make it through a boot, and then
> continue to function through reloads, as long as power is maintained.
> If power is turned off, even briefly, they go into a reboot cycles
> again. Is there a solution to this? (maybe question 4 below).
> The followingt error messages are displayed below. I don't think the
> checksome directly causes the reboot, but it may complicat things. The
> cycle for #1 is pretty quick, but that for #3 is very long, it will sit
> with the console idle for a LONG time (I got 164K of console log for #1
> but only 18K of console log for #3):
> #1 Exception: Software forced crash at 0x1385C6 (PC)
> #3 ENVM checksum (47 != 56) at offset 0x1
> #3 Local Timeout (control reg=0x110) Error, address: 0xFFFFFFFF at
> 0x2ADAA (PC)
Not sure about the second #3 but #1 and the first #3 look like software
corruption of some kind.
> 2) #3 does not recognize the ethernet card, even when I put in the
> ethernet board from the other machine that did work there, it also
> doesn't recognize one of the token ring cards. (see below, I think I
> know the answer to this)
The 6*ethernet card (CSC-MEC) is a cBus card and has to go in slots 5,
6, 8 or 9 with the CSC-CCTL2 board in slot 7. The MCI bus runs to all 9
slots, the cBus only the bottom 5. There are also two versions of the
cBus and if you've got really old kit (unlikely with CSC/4s) you have to
make sure that all the cards are compatible.
> 3) Machine #3 finally boots into rom based ios (ver 10.0). A reload
> command brings it up from flash (ver 11.0). Will jumpers fix this or
> is it booting from rom because it had to reboot several times?
Flash, eh - that might well be a CSC-MC+ card then. We got rid of our
AGS+s around 10.3 so I don't know when flash started to be supported or
how to use it on that type of system. We always used to boot from ROM
or by TFTP. What do the boot commands in the config say?
> 4) There are 3 (or is it 9?) batteries on the ENV board. Can I replace
> them, or connect a low voltage power supply to them such as a universal
> AC adapter made for portable consumer devices (I would pluf that into a
> ups), and will that solve my problem with failures/reboots when
> powering up?
ISTR it being 3 and them looking like standard AA NiCds. I have no idea
if they could be replaced or bypassed, but the point is that they're
there to preserve the NVRAM contents while the power is off - using a
UPS seems like overkill!
> ================================================== ======
> OK, I found that I have to configure the CCTL2 for the other boards in
> the ciscobus, that probably explains the boards that were not
> recognized.
You have to install the CCTL2 and the other boards in the right slots
but there's nothing else to configure on the CCTL2. (There are DIP
switches to set a "unit number" but I can't remember what they're for
and the manual I have (Hardware Products Installation Guide, October
1990) doesn't help, and anyway it only covers the CCTL not the CCTL2).
Sam