[zfs-discuss] zfs-discuss Digest, Vol 43, Issue 3

Gionatan Danti g.danti at assyoma.it
Wed Nov 7 15:55:24 EST 2018


Il 07-11-2018 19:52 Gordan Bobic via zfs-discuss ha scritto:
> This is a good paper from CERN on the matter:
> 
> http://indico.cern.ch/getFile.py/access?contribId=3&sessionId=0&resId=1&materialId=paper&confId=13797
> [6]
> 
> Their findings were that in reality the error rates are nearer 10^-7,
> IIRC.

Yeah, it is an interesting (olbeit old) read. It should be noted that 
10^-7 is the error rate of the *entire system*, not only from the disks 
themselves. From "RAID 5 verification":

"The second observation is the that we measured about 300 errors while 
from the
mentioned BER (Bit Error Rate) and the usage one would expect about 850
errors. As the vendor BER numbers are normally on the ‘safe’ side the
measurement is close to the expectation."

Regards.

-- 
Danti Gionatan
Supporto Tecnico
Assyoma S.r.l. - www.assyoma.it
email: g.danti at assyoma.it - info at assyoma.it
GPG public key ID: FF5F32A8


More information about the zfs-discuss mailing list