On Tue, Oct 16, 2001 at 02:07:13PM +0200, Posern wrote:
> But just for the future - cause I still like RAIDs - only this IBM
> harddisc modell "DTLA 307045" (or something like this)...
> Is total crap - In half a year I now have 5 broken harddiscs from this
These drives are known to be quite unreliable. See here for further links
> I have to check the /proc/mdstats to always see if all harddiscs are
> running or if the raid is in degraded mode (means that one harddisc
> is not in use - because of errors, or???) ???
Even better would be to have some sort of logwatching utility, or other
script which keeps an eye on it and emails you if something goes awry.
Alternately, you could have a hot spare drive sitting in the machine,
waiting to take over, without operator intervention, in case one of the
live drives goes south.
> What schould I had done at the point of the failure of the second disc
> to rescue my data???
> Because the raidhotadd thing is only possible if one harddisc fails!
If you have a standard RAID-5 setup, and you lose TWO drives, you're out
of luck. Nothing to do to rescue...
> To backup my raid-data nevertheless it is a raid5 - but what is a good,
> safe and fast method to backup 130GB of data???
> And - because it is my private raid and I am a poor student - a most
> likely a cheap method to backup...??!
Good, safe AND fast? You'd do fine with a DLT drive and a couple tapes.
But not on a student's budget. :^)
130GB is an awful lot of data, and it's not cheap to backup easily. Is
all 130GB actually *data* that can't be reloaded off a CD somewhere, or
recreated through some other method?
Also depends on what you need the backup for. Your RAID5 system should
protect you from drive failure (assuming you're aware of it when one drive
starts to go). But it's not going to protect you from an errant 'rm -rf'
command executed on the wrong directory.
Hope that helps...