On Fri, Sep 2, 2011 at 16:16, John Andersen
On 9/1/2011 11:38 PM, Per Jessen wrote:
Brian K. White wrote:>>> If you didn't happen to have any spare scsi disks on hand, most people
today would not be able to lay their hands on one in less than 24 to 48 hours no matter how much they're willing to pay for overnight,
If ordered before 17:00 and in stock, delivery usually follows the next day with regular Priority Mail. (that's from a normal web-shop available to everyone).
So posts the man from Zürich, a country of 15,940 square miles ;-)
Regardless, a hardware array will typically be configured with 1 spare drive sitting idle, so you'd be replacing the failed drive a while after your array had already recovered. No need to hurry at all.
Plus one on the hot spare(s).
Since they seldom looked at the server, and never rebooted it they got no warning of this at all.
If my Software Raid so much as hiccups, I get an email from mdadm.
If you buy all your disks for the array at the same time you have to be prepared for them to fail very close to each other.
FWIW: In my scenario it's a 1U server with only 2 drive bays. It took a week to get the replacement drive, but when the drive failed I got an email alert from Nagios, a free/open-source software to monitor servers. It's bad practice to setup a server and not put these measures in place. If you really value your data you shouldn't have some system in place to monitor it, be it you login every day and do the checks manually, or you implement something to do it for you on an automated basis. If you can't afford downtime, more important than what RAID you use, is how have you planned for when (not if) downtime happens? Wouldn't you feel silly if you setup the best RAID in the world and suffered a data loss because someone tripped over the extension cord between the server and the battery backup? -- Med Vennlig Hilsen, A. Helge Joakimsen -- To unsubscribe, e-mail: opensuse+unsubscribe@opensuse.org For additional commands, e-mail: opensuse+help@opensuse.org