[Date Prev][Date Next] [Thread Prev][Thread Next] [Date Index] [Thread Index]

Re: Poor failure recovery

Monday, October 11, 1999, 4:02:52 PM, John wrote:
> I want to know why we don't have any statically-linked essential
> utilities on a standard install.  Shell, cat, cp, ls, mv, rm, fsck,
> etc.

    Read the archives.  There was a big discussion over this just last month.

> Some might suggest "use the boot floppies."  Not always an option.
> Sometimes they don't contain versions compatible with the one on the
> installed system, meaning LOTS of extra work.  Besides, why introduce
> another failure mode (lost disk, bad floopy, bad floppy drive, etc) to 
> system recovery?

    My take, which I expressed then, is that there are two modes of operation.

    The first is where you have the time to use boot floppies. In this mode
there isn't a need because it is isn't mission critical to get the system back

    The second is where there isn't time.  If there isn't time, it is faster
to have spare machines laying around so you can fix the broken one at your
leisure.  In fact, if it really is mission critical then there are multiple
machines on the service so even if one fails another is there to take its

    To my way of thinking, if you are in a mission critical situation, your
machine fails, a boot floppy would take too long and you lose production
because of it then you need to rethink your mechanical-level architecture, not
the OS/software level.

    Others, obviously, disagree.

         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: