CentOS - Apr 2013 - DIMM problem

Hey, folks,

   I've got an HP Proliant DL580 G5 throwing ECC errors. This is annoying,
since a) it's all new as of a few months ago, and b) it's *fully*
populated. The two things I need to figure out are a) *which* DIMM it
is, and b) is it mirrored; if so, which *other* DIMM needs to come out
until we get replacements from the OEM.

Here's one of many, all identical, from dmesg:
EDAC MC0: CE row 12, channel 1, label "": Corrected error (Branch=0,
Channel 1),  DRAM-Bank=2 RD RAS=8218 CAS=500, CE Err=0x10000,
Syndrome=0x6cad8e02(Correctable Patrol Data ECC))

I see the Bank=2, so I assume that's the first riser board on the left;
but I can't identify which of the four (?) DIMMs on it is the problem.

I've been googling, and skimming useless manuals, and have just been
trying to look under /sys/devices/system/edac/mc/mc0/. I see ce_count
there showing thousands; but all of the ce_count files under csrow[0-7]
show zero.

Clues, anyone?

         mark

On 4/24/2013 10:34 AM, m.roth at 5-cent.us wrote:
> Hey, folks,
>
>     I've got an HP Proliant DL580 G5 throwing ECC errors. This is
annoying,
> since a) it's all new as of a few months ago, and b) it's *fully*
> populated. The two things I need to figure out are a) *which* DIMM it
> is, and b) is it mirrored; if so, which *other* DIMM needs to come out
> until we get replacements from the OEM.
>
> Here's one of many, all identical, from dmesg:
> EDAC MC0: CE row 12, channel 1, label "": Corrected error
(Branch=0,
> Channel 1),  DRAM-Bank=2 RD RAS=8218 CAS=500, CE Err=0x10000,
> Syndrome=0x6cad8e02(Correctable Patrol Data ECC))
call HP, that new server should be under support contract, no?


-- 
john r pierce                                      37N 122W
somewhere on the middle of the left coast

On Wed, 24 Apr 2013, m.roth at 5-cent.us wrote:
> Hey, folks,
>
>   I've got an HP Proliant DL580 G5 throwing ECC errors. This is
annoying,
> since a) it's all new as of a few months ago, and b) it's *fully*
> populated. The two things I need to figure out are a) *which* DIMM it
> is, and b) is it mirrored; if so, which *other* DIMM needs to come out
> until we get replacements from the OEM.
>
> Here's one of many, all identical, from dmesg:
> EDAC MC0: CE row 12, channel 1, label "": Corrected error
(Branch=0,
> Channel 1),  DRAM-Bank=2 RD RAS=8218 CAS=500, CE Err=0x10000,
> Syndrome=0x6cad8e02(Correctable Patrol Data ECC))
>
> I see the Bank=2, so I assume that's the first riser board on the left;
> but I can't identify which of the four (?) DIMMs on it is the problem.
>
> I've been googling, and skimming useless manuals, and have just been
> trying to look under /sys/devices/system/edac/mc/mc0/. I see ce_count
> there showing thousands; but all of the ce_count files under csrow[0-7]
> show zero.
>
> Clues, anyone?
Is there anything in the iml log on ILO? Also did you try just re-seating
the memory or moving it into other slots to see if you can track it down that
way??

Regards,

-- 
Tom			me at tdiehl.org		Spamtrap address	 		me123 at tdiehl.org

I recently switched our workstations from gdm to kdm, because I was 
receiving some complaints with gdm about having to set the desktop 
halfway through the login (versus being able to set it from the start of 
the login process), as well as having the mile long list of previous 
logins (which, by the way, if anyone knows how to turn that off, lemme 
know, please!).

After switching the window manager to kdm, we started having problems 
with our vncserver sessions. Every single session was kde, without 
regard for the contents of the xstartup file for that particular user. I 
have users that want gnome and others that want kde, and some project 
accounts that use some other custom stuff, so obviously this was a problem.

The only "configuration" I did after yum group-installing KDE-Desktop 
was to set up the /etc/sysconfig/desktop file to contain:
DESKTOP="KDE"
DISPLAYMANAGER="KDE"

so I am not sure what is up with kdm to be causing this. I tried looking 
online but only found some ancient debian/ubuntu questions that are 
mostly unanswered or resolved in ways that won't work for us (hard 
coding something in xinitrc, for example).

So, does anyone have any ideas as to what is going on and what I can do 
to resolve it? I am currently in the process of rolling back the 
workstations to the gdm configuration as a "fix", but if possible
I'd
really like to be able to actually sort out what is going on. Our 
systems are not heavily customized, being "software development 
workstation" installs with a few extra group installs (including the 
centos base tigervnc-server package) and a handful of third-party 
rpmforge packages installed, and kdm seems to work properly on the 
console (allowing selection of other desktops, etc), it's only with the 
~/.vnc/xstartup that we are having issues.

Thanks!
Miranda