-
Notifications
You must be signed in to change notification settings - Fork 0
null pointer exception in Mondrian #151
Comments
@hwchen can you share a link that replicates the issue or the link to a query where the issues originally occured? |
@jspeis unfortunately I don't know if the issue is replicable. Before flush, this query reliably gave an error: |
@hwchen would you be able to check the journalctl logs for both the |
from canon site logs, here's the trace:
|
Looks like some sort of internal Mondrian error? sigh @hwchen are there any other errors in the log around the same spots as these? |
The api and site logs differ a bit, possibly because of caching. So it's a little hard to line them up. The other error that happens reliably is this, and this will go away with flushing or just retrying sometimes.
|
@hwchen perhaps we can try making sure mondrian is getting passed these settings? https://help.pentaho.com/Documentation/5.3/0P0/100/050/010/040 not sure if it will make a difference but maybe? |
Weird, can't seem to find a University query associated with the Null pointer exception. I wonder if something else triggered an error in Mondrian, and then the the null pointer exception started showing up somewhere else? Anyways, here's the log for the cube, searching for
|
Interestingly, the errors that canon reports are only from October 3 onwards. |
@hwchen check this out pentaho/mondrian#1038 -- I think that's might be the same issue we're seeing. I wonder if there's a way we can try to apply that patch |
going to close this out as it seems the NullPointerException has not recurred! |
Ipeds completions cube developed a null pointer exception when drilling down on University dimension.
Trace was unhelpful, only noting that a null pointer exception occurred during generation of the sql query. (Unfortunately, I lost the exact trace, but it wasn't that helpful).
Flushing Mondrian fixed the problem.
The text was updated successfully, but these errors were encountered: