[ntp:questions] NTP errors are caused by single and multi-bit memory errors (was: Re: questions Digest, Vol 66, Issue 7)

David J Taylor david-taylor at blueyonder.co.uk.invalid
Tue Apr 13 06:07:00 UTC 2010


"Rob" <nomail at example.com> wrote in message 
news:slrnhs6ijb.i40.nomail at xs8.xs4all.nl...
> David J Taylor <david-taylor at blueyonder.co.uk.invalid> wrote:
>> "Todd Glassey" <tglassey at glassey.com> wrote in message
>> news:4BC31C47.1000606 at glassey.com...
>> []
>>> I want to know about non ECC memory based PC's running NTP and how 
>>> many
>>> NTP errors are caused by single and multi-bit memory errors.
>>>
>>> Todd
>>
>> I am not aware of any.  I rather imagine that memory errors in such a
>> system would likely cause the system to halt, rather than affecting 
>> just
>> NTP.
>
> On a non ECC system, single and multi-bit memory errors just go 
> undetected
> and do not cause any halt.  When they occur at an appreciable rate, 
> there will
> probably a lot of mysteriously crashing programs.  At still a higher 
> rate,
> the entire system will crash or hang.
>
> But there is no deterministic behaviour that you can analyze.

Yes, I stand corrected.  The "halt" would be because of some overwrite or 
consistency check failure by the OS.  I had in mind a hardware-detected 
error - obviously not with non-ECC systems.

What was behind Todd's original request?

Cheers,
David 




More information about the questions mailing list