I've tracked it down to just the # of queries / how busy the sever is.

Yet everything is in memory with 1 gig to spare.

I'd hate to goto a queue system to run queries, does anyone have any idea of
restricting the load on the server? Its not a timeout issue either. the
queries will run and return fast, its just when there are enough at once it
freaks out and restarts.

Anyone have similar problems before? Suggestions?


"PCH" <pch12 (AT) hotmail (DOT) com> wrote

Hello,
I had the similar problem twice:
when I backed up and restored my olap db from dev server to prd. Immediatly
after olap db have been restored and users try to query the cube, as starts
and then start to stop.
The solution was: delete all files in data folder of this database and then
reprocess all dimensions and cubes in the database.
The second time it was when I edited the cube, olap service sop with the
same message and try to start. As my database was enought big, olap service
starts about 1 hour.
The solution was: set olap service startup to manual, restart the server.
Then goto to the data folder and delete file "commit".
I thing universal solution should be delete all files in the data directory
and then reprocess all the cubes in the data directory.

Ramunas Balukonis


"PCH" <pch12 (AT) hotmail (DOT) com> wrote