hi all,

   My company is developing a network server/router for hotels (on Linux).
Essentially, guests in the hotels can connect
to the hotel's network with their laptops, and get to the Internet easily.
We provide billing and port management
and security features within our server.   

   One of our customers is very concerned about reliability, and I have
promised to investigate and learn what I
can about RAID and some of the options.  I've read the Howtos and so forth
(although I have had some poor luck with
some of the documents), and I am still unsure about some of my options.   I
am hoping that y'all can help me out.  I
don't need a lot of technical advice right now - I just need some opinions -
which of these options are viable for
a commercial, standalone, sitting-in-a-closet server, and which ones aren't.

Software RAID:

        I am concerned about two things - the possibility of hot-swap, and
boot-sector errors

        Boot Sector Errors:
                 if the power glitches,  for example, and the CPU starts to
write garbage to the drives, how well does
                the system recover - I have seen some comments on the
root-raid HOWTO, but nothing has clarified
                this for me - is this not an issue any longer?  I get this
feeling from the root-raid howto that I am
                expected to be standing by with a floppy in case it fails,
or have a floppy in the drive
                , but that isn't a good answer for this particular
implementation

        Hot-Swap and automatic rebuild
                The last document I read talked about Hot-Swap being
pre-alpha - but that document seemed to be
                out of date - has anything changed there?

                

Hardware RAID:

        I have two empty 5 1/4" bays in my system - if I get the proper RAID
SCSI controller, and hook it into two hard drives in removable chassis
        in those bays, is that a cheap hot-swap system - or do you need
three drives?   (I'm assuming RAID level 5 is the right level)

        Lastly, remote failure detection is very important - as I understand
it, these SCSI-to-SCSI controllers mostly just 'beep' when they have a
        failing drive.  Is it fundamentally impossible to discover that a
drive has failed via an audit, or is it just a matter of getting the right
driver
        interface?

Thanks a bunch,

        John



-----------
John Brothers   -       [EMAIL PROTECTED]              -       (678) 297
3084

Reply via email to