Re: itp: static bins / resolving static debian issues
Thursday, August 19, 1999, 4:53:52 AM, Marek wrote:
> Come on. You put up a brand new machine with a shiny, new Debian
> installation on it. Suddenly your 2GB SCSI drive fails for some reason.
A brand new machine wouldn't have anything on it, now would it?
> The system was brand new, was it a "known crippled server"? You're root
> partition on the broken disk is damaged, but your other disk which has /usr
> and /var/www on it still works. What you have to do? Hotswap the broken disk
> and copy a backup of the root dir and do it all without downing a server.
> With statics it's possible.
Since it isn't in use, it doesn't matter how long it takes or what methods
are used to bring it back to where it could be put into production.
> Yeah, give me those hundreds of thousands of bucks and I'll do it.
If you needed the service to be up, you would have the funding for
redundant hardware. Also, it would not be hundreds of thousands of dollars.
The class of machine that Debian typically runs on (x86) as a server is, at
most, a few thousand. <$10k and you could have some powerful machines running
redundant hardware.
The only time I can see redundant hardware running into hundreds of
thousands of dollars and have Debian on it is if you bought the top of the
line VArServer 3500 with a complete raid. However, anyone pushing that much
data around should be using a NetApp which, at that point, IIRC, costs less.
> One has to know more than one httpd around, boy. roxen doesn't fork new
> bins, use roxen if you're concerned.
I'm not concerned, Justin is, get it right.
> What is the reason to convince you?
I haven't thought of one yet. As I said, either you don't need to worry
about it, or if you do, there are other, better steps to take to ensure no
loss of service.
--
Steve C. Lamb | I'm your priest, I'm your shrink, I'm your
ICQ: 5107343 | main connection to the switchboard of souls.
-------------------------------+---------------------------------------------
Reply to: