[Mondrian] Pool.get()

Julian Hyde jhyde at pentaho.com
Tue May 19 16:19:33 EDT 2009


PS See also  <http://jira.pentaho.com/browse/MONDRIAN-496>
http://jira.pentaho.com/browse/MONDRIAN-496 and related issue
<http://jira.pentaho.com/browse/MONDRIAN-557>
http://jira.pentaho.com/browse/MONDRIAN-557. It's possibly that these are
causing extra work to be done at schema load time.


  _____  

From: Julian Hyde [mailto:jhyde at pentaho.com] 
Sent: Tuesday, May 19, 2009 1:00 PM
To: 'Mondrian developer mailing list'
Subject: RE: [Mondrian] Pool.get()


I don't know if there would be integrity issues if two users were using the
same URL but ended up with different schema objects. Obviously it would be
inefficient: each schema would have its own cache.
 
The ideal behavior would be for user B to wait if it needs the same schema
that A is currently loading. But if C comes along and wants a different
schema, there should be a very short critical section where the system
figures out that the schema does not exist and needs to be loaded.
 
Can you please log a bug for this issue?
 
Julian

-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://lists.pentaho.org/pipermail/mondrian/attachments/20090519/9be1f174/attachment.html 


More information about the Mondrian mailing list