[Info-vax] FREESPADRIFT

[Kerry Main]

> -----Original Message-----
> From: Info-vax [mailto:info-vax-bounces at info-vax.com] On Behalf Of
> Paul Sture via Info-vax
> Sent: 17-Jun-16 2:53 PM
> To: info-vax at info-vax.com
> Cc: Paul Sture <nospam at sture.ch>
> Subject: Re: [New Info-vax] FREESPADRIFT
> 
> On 2016-06-17, David Froble <davef at tsoft-inc.com> wrote:
> > Paul Sture wrote:
> 
> <snip>
> 
> >>
> >> Footnote: some concrete figures for money lost by extended
> downtime
> >> came out of that incident and management was not afraid to invest
> >> in beefing up their disaster recovery capabilities.
> >>
> >
> > Too bad that it appears to be a case of closing the barn doors after
> > the horses are out ....
> 
> Q: Why close the door after the horse has bolted?
> A: To stop the chickens coming home to roost!
> 
> >
> >:-)
> >
> > But, better late than never ....
> 
> Actually this was more a case of critical systems being disaster
> tolerant (DT versus DR). I don't think anyone had thought of the case
> where a whole building was out.
> 
> Initial solution: move selected systems to another building; just
across
> the road would do for starters, but leave you vulnerable to a street
> level failure.  Fortunately this customer was large enough to have
> existing data centres in different locations.
> 
> This is where the then recent arrival of Fibrechannel opened up
> possibilities - you could put 'local' disks in another building.
> 
> Longer term solution: build another data centre in another location
> and add a third data centre later.
> 

The dual active-active DC's within 100km of each other is the DR/DT
strategy that most Customers are looking at these days. For those that
need to accommodate a major incident taking out both of these sites,
then a third site is recommended much further from the first two i.e.
often called active-active-passive.

Why 100km? This is the rule of thumb distance which many folks in 
the industry state is a balance between site separation and yet still
allow sync writes to occur with acceptable performance. YMMV as
it all depends on things like the read-write ratio of the critical apps.

A-A-P is certainly more expensive, but DR/DT is like insurance - you 
may never use it, but if you do, you will be glad you made the 
investment.

What is usually not considered these days are pandemic issues like
SARS and [insert major pandemic]. In these days of global travel, it
Is a major concern for large Customers. A pandemic issue could
easily take out both "local" sites - all it takes is an OPS person with
something like SARS to visit both sites (perhaps picking up tapes?)
and inadvertently contaminate both sites. 

That's it - both sites are closed immediately once the guys in white 
suits get involved. Everyone goes home.

And in case anyone thinks this is pie-in-the-sky stuff, check out what 
happened to HP in Toronto during the SARS scare back In 2003 - luckily 
only 1 site was impacted: (the closed office included one major DC site)

http://www.cnet.com/news/sars-sends-hp-workers-home-in-canada/

http://www.itbusiness.ca/news/sars-scare-leaves-hp-canada-in-limbo/5669 


Regards,

Kerry Main
Kerry dot main at starkgaming dot com