[Samba] Re: Countless "rec_read bad magic" lines in log.smbd

[Charles Marcus]

On 5/8/2007 Michael Lueck (mlueck at lueckdatasystems.com) wrote:
> I have a diskette image of Memtest86, the previous version, so this 
> evening we will create another outage and try that. 

RAM that passes Memtest86 (even over a weekend of continuous testing) 
will not necessarily prove the RAM is good.

The only way I have been able to reliably test for bad RAM is to do 
something that stress-tests ALL of it - like compiling a kernel, or X, 
or something like that...

-- 

Best regards,

Charles