 |
 |
Thanks for the warning, I've taken out a trio of our adaptec nodes from our dpm and I'm draining them in preparation for running the RAL tests on 'em. Once we see the results of these tests we'll decide where to go next. We have 12 Adaptec raids in total... Cheers, Matt Ewan MacMahon wrote: >> -----Original Message----- >> From: Testbed Support for GridPP member institutes [mailto:TB- >> [log in to unmask]] On Behalf Of Matt Doidge >> >>> Have you run the RAL tests or anything like them on these machines, >>> and if not, is there any chance that you could (bearing in mind that >>> the RAL tests pretty much assume they can trash everything)? >> They received some medium-weight burning in but nothing as hardcore as the >> RAL tests (they were rushed into production as we'd been having trouble >> with our last batch of 3ware-driven raids). I could take one of them out >> of DPM, drain it and test it but that would take a little while. >> > If I were you I'd do it - we had/have three of the 36 bay versions in > our DPM and they were fine until one fell over last Monday in a manner > suspiciously similar to that in which the 24 bay does under the RAL > tests. I suspect that the load from ordinary DPM usage isn't heavy enough > to trip the problem very quickly/often. > > At the moment ours are all draining, then they'll get the RAL tests and > we'll see what happens. If they're going to die at intervals it's better > to find out sooner rather than later, and when they don't have too much > data on them. Draining (or trashing) 60+Tb of stuff wouldn't be fun. > > Ewan