Hello Leo,

"Braginsky, Leo" <[EMAIL PROTECTED]> writes:

> I am about to start evaluating Slide as content storage and retrieval facility. My 
>web application will be accessing Slide server programmatically to render pieces of 
>content on the screen. I am most interested in performance of content retrieval. Does 
>anyone possess any data that I can look at?
>  
> My conditions:
> 4) I was planning to use file system as opposed to JDBC to minimize complexity of 
>setup. Is this a good idea?

I don't think so. Recent slide requires a store with transaction support. This is not 
offered
by the file store. 

> 5) I will have to support a few thousand simultaneous run-time users accessing the 
>store. 

I can't give you any hard numbers.
We are running slide on Linux/tomcat/postgres and its working okay. The bottleneck for 
our applications is not content retrieval but postprocessing (we are running the 
content through several XSL stylesheets)
I have some doubts, if slide or even tomcat  can handle thousands of concurrent users, 
but it is worth a try. Of course it depends very much on the definition of concurrent.
 
> 6) Content is created off line and placed into Slide through WebDAV client (Internet 
>Explorer, for example). No performance requirements there. 

WebDAV support from IE sucks and Slide file store makes it worse. If you create a new
folder and rename it, there is some garbage left over in the file store, which makes
the folder unreadable. I could not reproduce this with JDBCStore. 

A more serious problem is, that IE does not forward any useful error messages to 
the user like 'Permission denied' etc. Of course you don't get any information about 
properties and so on.  This is not a Slide problem, but true for
all WebDAV Server. Use other clients for serious work, if possible. XMLSpy works
fine for us.


Martin

---------------------------------------------------------------------
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]

Reply via email to