Re: RAID 10

"Craig Kalugin" <CraigKalugin@xxxxxxxxxxxxxxxxxxxxxxxxx> wrote in message
news:0C0A1CA6-6DC9-414B-933A-F73EFC070A1D@xxxxxxxxxxxxxxxx
> What is the disaster if the controller failed in my RAID 10 config?
>

What is the disaster in case of a controller failure? In case of RAID1 or
RAID10 you are right the impact may not be as significant.

Well, if you are lucky and in mirroring scenarios you frequently are, the
data is not corrupt, you would replace the board, and the array would be
back online. A failed RAID5 controller will pretty much guarantee data
corruption, by the way. If you are not so lucky, you have striped mirrors,
and your stripe may not be possible to rebuild due to corruption. Yes, it is
rare, but I have dealt with enough failed RAID controllers to have a reason
not to trust them.

In either case, if a disk in RAID5 dies, users know. Duplex striping of
mirrors is one method I strongly prefer.

And those who claim that RAID10 is an unnecessary expense - my invoice for
recovery services will certainly make you rethink that.

In summary - RAID5 for file servers. RAID10 for database applications and
Exchange.

For SBS-load applications SATA controllers with NCQ are sufficient, so
expense should not be a concern.
--
Leonid S. Knyshov, CEO
Crashproof Solutions, LLC - http://www.crashproofsolutions.com
MCP Exchange 2003/Small Business Server 2003, CCNA, SCSA 8
Microsoft Small Business Specialist Partner


.

Relevant Pages

  • Re: 306GB drives!
    ... >> datacenter in which the EVA is located is destroyed by a disaster. ... > unless your redundant controller is in another building up to ... But more commonly in the OpenVMS Cluster world, ... at the remote site). ...
    (comp.os.vms)
  • [OT] 8-port AHCI SATA Controller?
    ... I'm looking for an 8-port SATA controller based on the AHCI chipset, ... On each SX8 and one onboard. ... handles one request per controller to safeguard against corruption. ... o It was also harder to hit with larger chunk sizes for the SW raid. ...
    (Linux-Kernel)
  • Re: Restore Active Directory from tape
    ... we already have a backup domain ... unless you have a disaster and everything ... This domain controller would be used in a disaster situation where ...
    (microsoft.public.windows.server.networking)
  • Re: 306GB drives!
    ... > datacenter in which the EVA is located is destroyed by a disaster. ... unless your redundant controller is in another building up to ... you have 4 controllers. ...
    (comp.os.vms)
  • Re: bad blocks on raid5 cause filesystem failure
    ... It is setup in in RAID 5. ... How could a RAID controller botch this up? ... should be rebuilt on spare drive if available from remaining drives. ... All is fine unless you have double fault. ...
    (comp.os.linux.hardware)