On Thu, 30 Sep 2010, Jonathan Nicol wrote: > On Sep 30, 2010, at 6:53 PM, da...@lang.hm wrote: > >> PSU's can die over time, and there can (and have been) flaws in >> design or components that will cause similar devices to start >> failing at around the same timeframe. >> >> If I start having them fail on several servers, I accelerate efforts >> to replace that generation of servers. >> >> the most common thing to fail in a PSU are the capacitors, and if >> the vendor had a bad batch of caps that made it into the power >> supplies, I would be reluctant to just replace the PSUs and put the >> systems back into mission-critical use, if those caps are bad, how >> can I really trust the other caps in the system? >> >> I am in the process of doing this with a batch of systems purchased >> 5-6 years ago. >> >> David Lang_______________________________________________ > > > Very true about them failing in batches, but.. the PSUs are almost > never made by the same manufacturer as the motherboard etc. We had a > batch of Sparkle PSUs all fail within months of each other, but the > Supermicro motherboards are still going strong years later..
forwhitebox vendors you are absolutly correct. I wouldn't be as comforatable making the same bet for companies that have their own boards made. And in the case of multi-PSU systems, the switching circuits that combine the PSUs is a potential problem, and that probably was made by the same company as the PSU. I ran into a batch of IBM RS6000 systems a couple years ago that had those boards act up on us. They started reporting that both power supplies had been turned off (taking the system down) David Lang _______________________________________________ Tech mailing list Tech@lopsa.org http://lopsa.org/cgi-bin/mailman/listinfo/tech This list provided by the League of Professional System Administrators http://lopsa.org/
