bri at sonicboom.org
Fri Sep 7 12:59:09 PDT 2001
raid 1 or 5 with a few disks, its really not that pricey now..
Brian "Sonic" Whalen
Success = Preparation + Opportunity
On Fri, 7 Sep 2001, Stephen Cope wrote:
> I just had a really nasty experience with a SCSI disk deciding to spin
> down and act silly.
> Anyway, its working all good for now, *but* in the event that something
> breaks later on, what should be done?
> At the moment my backup solution isn't too good. Grab a coupl'a dozen
> directories out of the web tree, tar it, and scp it over to another system.
> However in the event that the system does die, that backup doesn't help
> much because it doesn't include configuration files, specially compiled
> binaries, etc, etc.
> I considered rsync'ing / to a directory on a remote system.
> The situation is that the remote system is running ... Win98. Hmmm, no file
> permissions or ownerships. That can cause trouble when it gets restored.
> rsync has the advantage of *not* having to copy 3GB of data over an
> Ethernet connection (not Fast Ethernet, not Gigabit Ethernet).
> Does anyone have any proposed solutions?
> And also another thing ... how frequently should I back files up? I'm
> really conscious that between the time the last backup was made and when
> the system dies files will be lost. What's a way to minimise that?
> (Why am I conscious of it? Well, the current backup has been running for
> 20 minutes and mail has been arriving all during that time.)
> Hope this is the right place to ask ... especially since Chris has Tracy's
> "If you can't afford a backup ..." quote in his sig.
> Stephen Cope - http://sdc.org.nz/
More information about the KPLUG-List