ZFS errors on the array but not the disk.

[Russell L. Carter]

-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1



On 10/27/14 18:47, Zaphod Beeblebrox wrote:
> Well... why wouldn't this trigger an error with (say) the checksums
> on the devices themselves?  Without throwing an error, why is the
> vdev re - resilvering?  I don't have spare hardware to throw at it.
> It's otherwise a sane system.  It can "make -j32 buildworld"
> without choking.  It can download several hundred torrents at a
> time without corrupting them. Hardly seems like suspect hardware.

I will just say as a non-zfs expert that I have had several disastrous
raid failures over the last 15 yrs, and a couple that cost me real
money, and it was always hw.

And the reason it was disastrous was I couldn't diagnose it even though
I was a pseudo-expert.  I spent a lot of time under deadlines assuming
the underlying hw was sane.  The software diagnostics were no help.
I trusted the hw then, but no more.  And your reports (thank you) are
a reminder to me to not give too much credence to zpool status.

BR,
Russell


-----BEGIN PGP SIGNATURE-----
Version: GnuPG v2
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=SOhS
-----END PGP SIGNATURE-----