Re: Type of server that would be used

I should update that documentation - the ROADMAP lays out our
development plans to be the gradual shift to postgres/apache.

At this point, I'm about dead set against letting the python based
daemon live, but the shelve implementation will probably stay.

IE, if you want high end all features on, then sql + apache is the way.
if you want a really scalable anonymous server, then shelves + apache
would be very nice.




On Wed, 2002-04-03 at 14:20, Bingner Sam J Contractor PACAF CSS/SCHE
wrote:
> >* Another ongoing question is how valuable people will think a "standalone"
> 
> >  server will be, versus a fast/featureful server. 
> >  
> >  IE, right now to get up and running, you unpack current, run stunnel,
> >  and go. But its slow, since we're doing http in python. We're also 
> >  using python shelves for all our data storage, which limits how useful
> >  client data storage can be (shelves are simple key:value pairs really,
> >  and you need more power to support the really advanced features).
> >  
> >  We could increase power by using apache to do the heavy bit shifting, 
> >  and postgresql to be our main storage of data. But that means that 
> >  a new site would have to get apache setup and running, get postgresql
> >  setup and running, and then tackle current.
> >  
> >  Sure, having both would be great. I'll do that if I can figure out how,
> >  but would anybody really use one or the other? 
> >
> 
> 
> 
> I know that I would much prefer to have an SQL backend and a webserver.
> Performance and reliability are much more important to me than simplicity of
> initial installation, or valid pre-requisites.  I personally feel that if a
> Systems Admin who is trying to set up an up2date server to support all his
> systems, who cannot figure out how to install and configure Apache and some
> sory of SQL Database really has no business being a Systems Admin.
> 
> Joe Blow home user is never going to want to run current unless they're just
> intrigued by the posibility and are interested in learning.  If that is the
> case, they are not going to mind installing a little extra software
> either... and would probably enjoy the challenge.
> 
> That said, we would of course have very nice and simple to follow
> documentation on how to install and configure all relevant services... so
> somebody SHOULD be able to just walk through that anyways.... right?
> 
> 
> Sam Bingner
> _______________________________________________
> Current-server mailing list
> [email protected]
> http://lists.dulug.duke.edu/mailman/listinfo/current-server
> 
> 
-- 
Hunter Matthews                          Unix / Network Administrator
Office: BioScience 145/244               Duke Univ. Biology Department
Key: F0F88438 / FFB5 34C0 B350 99A4 BB02  9779 A5DB 8B09 F0F8 8438
Never take candy from strangers. Especially on the internet.