11G wrong IPMI data

[Peter Kjellstrom]

On Friday 10 July 2009, Patrick Schreurs wrote:
> Peter Kjellstrom wrote:
> > On Friday 10 July 2009, Patrick Schreurs wrote:
...
> >> ~# ipmitool sdr|grep cr
> >> Temp             | 53 degrees C      | cr
> >
> > Do you have a 2nd R710 so that you can see if this is "normal".
>
> Yes, we have several. They all report the same thing.
>
> > I'd also try running OMSA to see what it reports. IPMI is usually a lot
> > less tested on servers (unfortunately...).
>
> omsa doesn't see a problem at all:
>
> Index                     : 0
> Status                    : Ok
> Probe Name                : System Board Ambient Temp
> Reading                   : 18.0 C

But isn't that two different sensors? (the one IPMI deems critical and the one 
OMSA calls "System Board Ambient Temp"?)

...
> > I've seen many servers that simply report (sometimes very) wrong
> > temperature figures via ipmi.
>
> Our monitoring relies on ipmi, which make this very annoying.

Ours too so yes I'm with you on this :-) I'd prefer the solution "Every server 
delivered with well behaving IPMI".

But, in reality, to get this to work one basically has to select one or a few 
sensors that look good and then monitor those only (with possible offsets). 
Maybe you can find one sensor that seems to move nicely with the server 
temperature (or some definition thereof) and simply consider "10 deg above 
normal" as a trigger for your monitoring.

Stuck in this imperfect world,
 Peter
-------------- next part --------------
A non-text attachment was scrubbed...
Name: not available
Type: application/pgp-signature
Size: 189 bytes
Desc: This is a digitally signed message part.
Url : http://lists.us.dell.com/pipermail/linux-poweredge/attachments/20090710/b733e489/attachment.sig