This thread was started on the zfs list and I guess would properly
belong here.  Below is a short summary of the thread  and a new
response.

,----
| From: Ian Collins <i...@ianshome.com>
| Message-ID: <4cb29bf8.8000...@ianshome.com>
| References: <87bp71e59o....@newsguy.com>
`----

On 10/11/10 05:13 PM, Harry Putnam wrote:
> Osol b134
>
> I'm experiencing a forced shutdown of a machine.
> It is proceeded with a number of beeps in a steady pattern like
>
>    beep beep           beep beep           beep beep
>
> And onward.  2 beeps pause 2 beeps pause... etc.
>
> The beeps are the same tone on each set of beeps but each has its own
> tone.  So its .. hi low pause hi low pause .. etc
>
> I'm not sure what it means but it appears to be somewhat serious since
> it will shutdown the machine at some point... and not too long about
> it.
>
> I suspect some fairly serious hardware problem.  The machine is
> somewhat old 7-8 yrs and is:
> athlon64 +3400 (cpu (might be) 2.2Ghz or so)
>
> It has 3 sets of mirrored discs
>
> 2 Sets of 500gb (WD)
> 1 set  of 750gb (WD)
>
> I'm running the machine on a KVM and had the opensolaris machine off
> the switch while having a laptop sharing the KVM.  I've been using ssh
> to login to the problem machine up until now... and now no ssh comes
> up when I boot it.
>
> I know I will have to rerig things and connect mouse, keyboard,
> monitor direct to the problem machine, but once that is done is there
> some sort of semi-standard way of booting to an emergency set of tools
> to try to see what the problem is?
>
> I won't have a lot of time, before the machine shuts down.  Maybe 1/2
> hr or so.
>
> I'm kind of having a brain freeze about this:
> So what are the standard tests or cmds to run to collect enough data
> to try to make a determination of what the problem is?   

Ian responded:
Your motherboard manual would be the best place to look, it sounds like 
the BIOS reporting something (failed fan?)

Harry wrote:
> And if worse case turns out to be the case and I have to get a new
> machine or some new hardware, can anyone offer a small outline of the
> steps that would be needed to move those discs to a new setup?
>
>    

Ian responded:
Plug them in.

You will probably have to boot form a live CD, export and re-import the 
root pool to get it to boot.

-------        ---------       ---=---       ---------      -------- 

Edward Ned Harvey wrote:
Edward Ned Harvey wrote:
Date: Mon, 11 Oct 2010 07:33:28 -0400
From: "Edward Ned Harvey" <sh...@nedharvey.com>
Message-ID: <000001cb6938$23f5afa0$6be10e...@com>
References: <87bp71e59o....@newsguy.com>

> From: zfs-discuss-boun...@opensolaris.org [mailto:zfs-discuss-
> boun...@opensolaris.org] On Behalf Of Harry Putnam
> 
>   beep beep           beep beep           beep beep
> 
> I'm kind of having a brain freeze about this:
> So what are the standard tests or cmds to run to collect enough data
> to try to make a determination of what the problem is?

Definitely hardware.

To diagnose hardware, no standard test.  Start replacing hardware.  You'll
know you fixed it when the problem stops.

[...]

-------        ---------       ---=---       ---------      -------- 

Harrys' most recent response:

I was able to get logged in for a while today and grabbed the last
logs from Oct 3.

I see piles of warnings like this:

Oct  3 17:36:01 zfs genunix: [ID 667643 kern.warning] WARNING: Postattach 
failed for hid8
Oct  3 17:36:01 zfs genunix: [ID 667643 kern.warning] WARNING: Postattach 
failed for hid17

But those are peripheral kinds of usb ports or such aren't they?

Where to look for any clues ... so far its been running for about 1/2
hr and not getting any beeps yet, so may be a good opportunity to
extract some kind of diagnostics

I can't get X up ... it just went to a black screen, after seeing the
main login screen, logging in to consol and calling:

but am able to
login at the terminal or ssh from elsewhere to the osol box.

I've posted dmesg at:

  www.jtan.com/~reader/dmesg_osol.txt

I haven't scanned it thoroughly but didn't see anthing I recognized as
a clue. 

If the thing was getting hot, would there be a record left somewhere?


_______________________________________________
opensolaris-discuss mailing list
opensolaris-discuss@opensolaris.org

Reply via email to