[CentOS] HDD badblocks

Chris Murphy lists at colorremedies.com
Tue Jan 19 22:39:30 UTC 2016


On Tue, Jan 19, 2016, 3:30 PM  <m.roth at 5-cent.us> wrote:

> Chris Murphy wrote:
> > On Mon, Jan 18, 2016, 4:39 AM Alessandro Baggi
> > <alessandro.baggi at gmail.com>
> > wrote:
> >> Il 18/01/2016 12:09, Chris Murphy ha scritto:
> >> > What is the result for each drive?
> >> >
> >> > smartctl -l scterc <dev>
> >> >
> >> SCT Error Recovery Control command not supported
> >>
> > The drive is disqualified unless your usecase can tolerate the possibly
> > very high error recovery time for these drives.
> >
> > Do a search for Red Hat documentation on the SCSI Command Timer. By
> > default
> > this is 30 seconds. You'll have to raise this to 120 out maybe even 180
> > depending on the maximum time the drive attempts to recover. The SCSI
> > Command Timer is a kernel seeing per block device. Basically it's giving
> > up, and resetting the link to drive because while the drive is in deep
> > recovery it doesn't respond to anything.
> >
> Replace the drive. Yesterday.


That's just masking the problem, his setup will still be misconfigured for
RAID.

It's a 512e AF drive? If so, the bad sector count is inflated by 8. In
reality less than 15 sectors are bad. And none have been reallocated due to
misconfiguration.


Chris Murphy



More information about the CentOS mailing list