Bad memory

Everything about the project RNA World
Nachricht
Autor
iconized

Bad memory

#1 Ungelesener Beitrag von iconized » 15.11.2011 01:01

Hi,

I recently discovered that one of my memory modules malfunctioned. Basically all memory in the range 2-4 GB and 6-8 GB was bad, using 4 x 4 GB in dual channel mode. The current problem is, that I can't be sure when it started to malfunction.

I have ran RNA World work units from July 31 till September 30. RNA World work units use lots of memory so we can be pretty sure it has used some of the (later?) defunct memory ranges. My question, is it likely that RNA World work units can finish successfully when the memory behaves badly? I would expect the application to crash though.

Regards, iconized DPC

Ananas
WU-Schieber
WU-Schieber
Beiträge: 1184
Registriert: 27.04.2008 18:37
Wohnort: Nordlichter Köln

Re: Bad memory

#2 Ungelesener Beitrag von Ananas » 15.11.2011 01:32

As long as the results validate properly, they have not been affected.

If the bad RAM modules had affected the active program code itself, that would have made the application crash.

If they affect the data region of the program (randomly flipping bits or something similar), the application still might run to success but the result will never validate.
vi BOINC/checkin_notes
:1,$s/bug/feature/g
:wq!

Erzaehlen sich Biologen eigentlich Klein-RNA-Witze?

Antworten

Zurück zu „RNA World Discussions (english)“