Re: ZODB and Web2py

On Sat, Mar 11, 2017 at 9:03 AM, Toby Shepard <[email protected]> wrote:

>
>
> To put a fine point on it, you'd be using a ZODB database in any case. The
> question is whether you use a client-server *storage* like ZEO,
> RelStorage or NEO, or use FileStorage.
>
> You can use FileStorage if you have a *single process*, but that's
> usually not a good idea because then you wouldn't be able to make a
> read-write connection to the database from outside your web application
> without stopping it.  You also wouldn't be able to increase the number of
> application processes for scale or higher availability.
>
> I wasn't sure whether web2py ran as a single process.  Does it fork to
> handle concurrent requests?
>

I wouldn't know. :)


> I was just looking for the easiest path.  It sounds like I need to have a
> client-server storage.   I've just
> begun testing ZODB with FileStorage.  Would I just drop in a client-server
> capable one in its place,
> or is there more to it?
>

Aside from running the server, and having a different storage config,
that's all there is to it.



> Which would be a good choice for this application, running behind web2py
> as a web application database?
>

That depends on your needs.  Do you care what server you run?

Have you looked at Newt DB? http://newtdb.org

With Newt, you're using RelStorage with Postgres, but you a secondary JSON
representation that you can see, index, and search in Postgres and you get
an API that takes PostgreSQL search criteria and returns database objects.

Jim

-- 
Jim Fulton
http://jimfulton.info

-- 
You received this message because you are subscribed to the Google Groups "zodb" group.
To unsubscribe from this group and stop receiving emails from it, send an email to [email protected].
For more options, visit https://groups.google.com/d/optout.