On 3/10/19 3:04 PM, Reindl Harald wrote:
the original meaning of RAID was "redundant array of inexpensive disks"
Redundant Array of INDEPENDENT Disks.
The consensus is that "inexpensive" came later, seeing as the early
disks were ALL expensive.
https://en.wikipedia.org/wiki/RAID
RAID (Redundant Array of Independent Disks, originally Redundant Array
of Inexpensive Disks) is a data storage virtualization technology that
combines multiple physical disk drive components into one or more
logical units for the purposes of data redundancy, performance
improvement, or both
Yes, Reindl, you are correct on this. It was "inexpensive" until drive
manufacturers sabotaged it, around 2011. Basically by removing TLER /
SCTERC from their desktop drive firmware, leaving the *unconfigurable*
extended (2-3 minute) error retry behaviour. Prior to this sabotage,
desktop drives would power up with the long retry, but could be set to a
short timeout, making them interface-equivalent to enterprise drives.
You are wrong about the seriousness of the problem modern inexpensive
(non-raid) drives present both to linux and to hardware raid
controllers. FWIW, if a drive claims to support raid1, it has proper
timeouts and is suitable for any raid level in linux.
Phil
