On 2016-05-19 17:37, Ted Creedon wrote: > During a drive upgrade a brand new WD29EFRX 2TB drive containing my RW > volumes crashed.
Yeah, HDs tend to fail at the beginning and at the end of their lifespan. So do a load test on the fist few days and if it runs well, it will run well the next few year for a high probability. Our drives runs fine for >7 years now - only 3 dropouts out of 60 HDs yet. Luckily we are able to replace them soon (tm), as 7 years is far over the usual lifespan of a HD. > After talking to drive recovery firms I've learned that the new reliability > philosophy is cheap drives & rely on the users RAID for recovery, Has always been that way, rely on RAID and backup, expect EVERY HD to fail every time. > Fortunately I have 4 RO backups, one is enough, in case of fire^^ > One repair firm said they see new drives with incredible amounts of dirt > inside. Depends on company, label,... > Here is a drive reliability report. > > http://www.extremetech.com/computing/228497-backblaze-releases-billion-hour-hard-drive-reliability-report This is a bit half way, as it does not really show at which age the HDs died, so those values are not really much useful. Also some drives tend to be bad only in some production cycles,... > Looks like solid state drives + thumb drives for off site storage is the way > ti go. If you got the money... We go with 8TB drives and a RAID6 across them. Enough capacity and speed for 20 OpenAFS clients and some VMs. > tedc MfG, Lars Schimmer -- ------------------------------------------------------------- TU Graz, Institut für ComputerGraphik & WissensVisualisierung Tel: +43 316 873-5405 E-Mail: l.schim...@cgv.tugraz.at Fax: +43 316 873-5402 PGP-Key-ID: 0x4A9B1723
signature.asc
Description: OpenPGP digital signature
