Re: continuous reboots in a two nodes cluster with heartbeat and pacemaker.
On 19 August 2012 02:32, Stan Hoeppner <stan@hardwarefreak.com> wrote:
> On 8/18/2012 6:36 AM, Mauro wrote:
>
>> I've upgraded ram from 32 to 64G.
>
> Did the reboots occur before doing this?
>
>> I've reinstalled all simms.
>
> DIMMs. SIMMs haven't been used for over a decade. But the fact you
> mentioned SIMMs tells me you've been at this game a while.
>
>> The bios reports no ram problems.
>
> It may not.
>
>> Also other server are upgraded to 64G.
>
> Are all the DIMMs fully buffered ECC DDR2-667? If you added unbuffered
> ECC DDR2-667 DIMMs to go from 32GB to 64GB, the BIOS may no throw any
> errors, but you'd see things like random reboots, random lockups, kernel
> errors, processes crashing, etc.
>
> Do these two machines have all four processors and all four memory
> cartridges installed, or only two of each?
>
>> Bugged ram is on both servers? Strange to me.
>
> It's probably not bad RAM, but it seems possible that you may have used
> unregistered DIMMs when you upgraded to 64GB. Again, the BIOS may not
> balk at this, and the combo might seem to work until the memory
> subsystem is sufficiently exercised.
Thank you for your replies.
I'll check the DIMMs and tell you.
Reply to: