At 23:23 31/03/2010, you wrote:
>On Wed, 2010-03-31 at 22:01 +0100, Mary Hawking wrote:
>
> > EMIS Web holds the mirrored records on two mirrored servers in Leeds -
> > and, as an EMIS user, I trust the company.
>
>
>Hmm.
>
>
>How far apart are the 2 servers?
>
>As an EMIS user, if I were an EMIS customer, I'd ask that, but I'm not,
>like all English GPs.
>
>
>I'd still wonder whether the two centres (assuming the 2 servers are
>actually in two different buildings, and not in wings of the same, as
>happened outside the oil storage depot's perimeter fence a few years
>ago) have two separate, independent, geographically divergent routes
>connecting them to Exeter.
>
>Within Exeter I'm perfectly well aware that I don't have two independent
>routes out, although I do have two buildings, and thus a quick move of
>machine could improve matters in some rare cirumstances, despite EMIS
>and the PCT frustrating the sensible[1] solution of using remote
>desktops on our Windows 2003TS, but that is another problem, and more
>like my problem.
>
>
>Thought experiment time: What is the diameter of the smallest circle of
>total destruction that I could notionally place on a map anywhere
>outside the city limits of the city my practice is in, and cause EMIS'
>remote hosted system to cease to function in my building?
>
>That seems to me to be a useful and sensible metric - if it is 100
>kilometres then we are into asteroid strike territory, if it is 10
>metres then we are still in territory where the sensible person would
>not go. For 100, 1000, 10000 metres there is range for discussion,
>between blown fuse, rogue plumber, riot and deliberate action.
Strangely enough I was watching (recorded) The Gadget Show earlier
this evening. Mainly nonsense, but fun in parts. However they
report that it is "advised" that backups are stored at least a Jumbo
wingspan away from whatever is being backed up, and identified that
as 60m (web suggests nearer 65m for the 747-400).
I'm not suggesting this is particularly more "Logical" than any other
proposal (unless you're based at an airport), but it does suggest
there may be a de facto standard somewhere.
I don't know how much damage is done (if any) to cables by a
lightening strike but that is a regular occurrence that should be considered.
JB
|