From mboxrd@z Thu Jan 1 00:00:00 1970 From: Brendan Conoboy Subject: Re: "raid" versions of hard drives for software raid? Date: Wed, 02 Jul 2008 15:30:24 -0600 Message-ID: <486BF370.2070401@redhat.com> References: <873amsqbag.fsf@uwo.ca> <486BE720.3090507@redhat.com> Mime-Version: 1.0 Content-Type: text/plain; charset=ISO-8859-1; format=flowed Content-Transfer-Encoding: 7bit Return-path: In-Reply-To: Sender: linux-raid-owner@vger.kernel.org To: Justin Piszcz Cc: Dan Christensen , linux-raid@vger.kernel.org List-Id: linux-raid.ids Justin Piszcz wrote: > That's a lot of RMA's, bad PSU or vibration issues, or a bad lot of drives? My educated case is that it was a bad batch of drives (Bought them sooner after their introduction). Drives in the enclosure from other MFR's have held up reliably and heat has been a non-issue. Dan might want to see if hdparm has a tunable for the TLER setting he's concerned about. Modern mdadm versions will attempt to rewrite a sector if there's a read error using parity or mirrors to reconstruct the data. For this purpose having a short timeout before getting an error would be preferable to waiting a couple minutes. -- Brendan Conoboy / Red Hat, Inc. / blc@redhat.com