 | | | Good documentation, particularly re: clustering | Good documentation, particularly re: clustering 2003-11-24 - By John Haxby
Back
Rob Kirkbride wrote:
> Having just downloaded the ISOs for 3 (we were previously on 2.1) I 've
> not yet looked at the documentation CDs that come with it as I 've not
> got these down yet.
> I 'd be grateful for any thoughts/experience/URLs, particularly from
> people who 've actually done it! It 's probably a little early for any
> books to have been published.
I read the AS2.1 documentation on the clustering and HA stuff and it 's
pretty good -- explains both the concepts and what you have to do. I 'd
say it 's a good start.
High Availablility (in general), is pretty esoteric. You start from
the premise that you want to be able to cope with failure of any one
component (a machine, a disk, a network, a power supply) and then set
your systems up so that (a) they 'll cope with the one failure and (b)
not trigger the failover stuff on a false positive. Sounds easy in
theory. In practice, there 's a huge amount to consider. For
example, if you share some disks between two machines by having SCSI
controllers on each machine then you need to be sure that your SCSI
controller will behave nicely when it doesn 't have any power -- cheap
ones don 't, expensive ones do. For a real HA set up, you need someone
in who understands _all_ this stuff because if you 're paying all that
money to get 99.999% (five nines -- five minutes unplanned downtime per
year) you want to make sure that 's what you 've got.
jch
|
|
|