[Info-vax] RAM errors?

[MG]

On 13-2-2011 16:00, MG wrote:
> Recently I had a very unfortunate problem with one of my rx2600s, which
> I had powered off for several days. Yesterday I powered it up again and
> rebooted it more than once. At some point, the system hung (after
> successfully booted into OpenVMS) and I believe I had to do a power
> cycle (via iLO). Later the EFI presented me with a series of memory
> (RAM) errors[.]

[...]

> After reseating the DIMMs on the above mentioned locations, the system
> 'works' again (up to this very moment of writing this message) and the
> errors have seemingly disappeared.

But I'm not very confident about how stable this will be, as to how long
this may 'last'...  (I probably should have emphasized that more.)


> The reason I rebooted was because sometimes there would be visual
> artifacts in the screen (in weird varying 'quantities'). Strangely
> enough, these visual artifacts appear to be caused whenever I'd have a
> simultaneous serial console connection or after I rebooted from a
> console. However, now it almost seems as if the visual artifacts are
> showing up more frequently than before.
>
> The video card in question is a Radeon 7500 (PCI), I have two of them
> and both are installed in rx2600s (they should be identical, I purchased
> them from one and the same retailer). Another thing that I've noticed is
> that the Radeon 7500 in the other rx2600 provides EFI graphics output,
> while the other one --- the system with the [memory] issues --- does
> not. Both rx2600s have the same firmware and I'm not aware of any
> specific Radeon 7500 firmware.

Maybe I should've split this into a seperate thread, but it's still
very much related (the fact that in one rx2600 it provides graphical EFI
output and in the other it doesn't, with identical card).


> Could all of this be caused by faulty memory, or is perhaps the PCI card
> backplane faulty? Maybe even the system board? Because, although the
> system is up and running again, I'm not very confident about it and I
> wonder for how long that will be...

I'm a student and OpenVMS Hobbyist with a limited budget, but since
I've invested this much already in IA-64 hardware and the fact that I'm
really interested and serious about OpenVMS, I wouldn't mind investing
in a replacement system board.  But, I wonder, will that solve my
problems?  In that light, it will be a costly bet.  Is it perhaps a
[processor] cooling issue?

I have been asking about this (and the EFI/BMC issue) everywhere, but
with zero responses so far.  I sure hope that this doesn't mean that
these problems are 'unique,' by any means.

Any help will be greatly appreciated.  Thanks in advance.

  - MG