[Mondrian] mondrian.rolap.RolapMemberBase cannot be cast tomondrian.rolap.RolapCubeMember
Martijn van den Broek
mjbroek at gmail.com
Thu Nov 18 11:12:09 EST 2010
Yeah first thing we tried. It seems the same message indeed. However we have
either explicit or default set the
mondrian.rolap.EnableRolapCubeMemberCache=true in properties.
Is there anything else we can try? How can I give more debug information?
I'm still trying to get a testcase together that demonstrates the issue.
On Thu, Nov 18, 2010 at 4:55 PM, Julian Hyde <jhyde at pentaho.com> wrote:
> Did you search jira for this issue? There are similar ones, e.g.
> *From:* mondrian-bounces at pentaho.org [mailto:mondrian-bounces at pentaho.org]
> *On Behalf Of *Martijn van den Broek
> *Sent:* Thursday, November 18, 2010 12:49 AM
> *To:* mondrian at pentaho.org
> *Subject:* [Mondrian] mondrian.rolap.RolapMemberBase cannot be cast
> We are running a mondrian server with multiple datasources configured. The
> XMLA protocol is used to execute MDX queries against mondrian with a simple
> HTTP client written in python. This all works really well till at some point
> the mondrian server wont execute anymore and returns with the following
> mondrian.rolap.RolapMemberBase cannot be cast to
> mondrian.rolap.RolapCubeMember (code: 00HSBE03)
> From that point on we cannot execute the same MDX over XMLA till we clear
> the mondrian cache and restart the tomcat server. While XMLA is still broken
> we can however execute the MDX successfully in JPivot. Does this mean its a
> XMLA problem? Another thing we noticed is that some dimensions are not
> accessable anymore if we execute MDSCHEMA_HIERARCHIES.
> The WEB-INF/mondrian.properties has the default config and changing the
> 'mondrian.rolap.EnableRolapCubeMemberCache' to true or false doesnt change
> the behaviour.
> We are working on a testcase to reproduce the problem but it seems very
> hard to trigger. Any suggestions on how to debug this issue are welcome.
> Thanks in advance.
> Martijn van den Broek
> Mondrian mailing list
> Mondrian at pentaho.org
Martijn van den Broek
-------------- next part --------------
An HTML attachment was scrubbed...
More information about the Mondrian