Re: Local/remote data synch

From: paul brown (pbrown74_at_hotmail.com)
Date: 03/10/04 

Date: Wed, 10 Mar 2004 00:35:47 GMT

Hi Davor,

If you are going to look at keeping objects in sync across processes then
you may
want to look at some sort of clustering between instances of your
application.

I've just started some open source stuff that you might be interested in. At
the moment
I only cluster Properties objects across instances of a server but I am
going to generalise
it in the next couple of weeks.

http://sourceforge.net/projects/rformgen

(formgen-0.2-full.zip is what you need to download)

FormGen also provides you with a form that allows you to get/set on the
clustered
objects at runtime through a dynamically generated web page.

If you have specific requirements then email me offline and we can discuss
them (details
in the download).

Hope this helps,
Paul

"Davor Cengija" <dcengija_IQ_Filter@inet.hr> wrote in message
news:c2jrro$qdu$1@sunce.iskon.hr...
> nos wrote:
> > I read a book on oracle and they discuss it there.
>
> Which book, if you remember?
>
> >
> > "Davor Cengija" <dcengija_IQ_Filter@inet.hr> wrote in message
> > news:c2hvi0$jtj$1@sunce.iskon.hr...
> >> This problems bothers me for some time now, and probably other
> >> people as well, but for my surprise, I couldn't find any meaningful
> >> pattern, description, case study or whatever.
> >>
> >> So... how do you folks synchronize the data between many local
> >> applications and server, so that at some point in time every local
> >> application has the complete data, regardless of where the
> >> particular detail was inserted?
> >>
> >> This is my situation: Server serves as a repository and has the
> >> referential data. Each client works independently and several times
> >> a day sends it's changes to the server and collects the changes
> >> produced by other clients. The synchronization algorithm is rather
> >> complex, including 'fetched', 'changed' and other flags for each
> >> record, as well as data diffs (in the form of SQL UPDATE actions
> >> which are executed on each synch). The system works but looks very
> >> architecturally unstable.
> >>
> >> I'm thinking of implementing the 'object synchronization' scheme,
> >> instead of this SQL one, but it's not yet clear what would object
> >> synchronization mean and do. :-)
> >>
> >> What do you, people, have to say regarding this issue?
> >>
> >> Regards,
> >> Davor
> >>
> >> --
> >> Davor Cengija, dcengija_IQ_Filter_@inet.hr
>
> --
> Davor Cengija, dcengija_IQ_Filter_@inet.hr
>
>

Relevant Pages

  • Re: Storport or scsiport?
    ... I know when I download a new version ... from our vendor, they clearly list both versions and I always take the ... MVP - Windows Server - Clustering ...
    (microsoft.public.windows.server.clustering)
  • Re: Fuzzy k means
    ... Or go here directly on FEX (fuzzy k means clustering) to download the file you need: ...
    (comp.soft-sys.matlab)
  • Re: How to Analyse for Lumpiness? (Correction)
    ... There must be some sort of analysis to uncover evidence of clustering. ... I seem to remember the guys who audited the UK Lotto did some sort of cluster analysis so I'll try to track them down. ... So far I've had one reply - a suggestion that I compare my closest triple from each Lotto draw with the average closest triple from the 14M 6/49 combos - of course, ...
    (rec.gambling.lottery)
  • Re: vn(4) performance on 4.11 versus md(4) on 5.4
    ... production -- the sync writes are a hack to work around hangs, ... native ffs gave 54MB/sec by mostly writing with a physical block size ... The size of 64K results from clustering ... deadlock problems were supposed to be fixed, ...
    (freebsd-performance)
  • Re: arp problem in /var/log/messages
    ... > looks like someone is using tools like ettercap. ... It could either be that - ARP cache poisoning - or some sort of clustering ...
    (FreeBSD-Security)