on 06-01-2006 3:32 AM
Hi,
I am running a query with about 16000 records, the dimension table has almost same amount. I got really slow when it is formatting the excel sheet. It almost take 20 minites, the report is very basic report, no calculation,no condition/exception, is that normal? any idea?
Thanks
Hi,
Use Toce RSRT with option 'Execute + Debug' and also select Display Statastics data.
Now check the value for QTIMEDB.If is taking alomost same time of how much time taken by query to execute, then you need to work to increase the performance by Indexing on that Dimension and Define that as Line item dimension.
With rgds,
Anil Kuamr Sharma .P
You must be a registered user to add a comment. If you've already registered, sign in. Otherwise, register and sign in.
Richard,
Welcome to SDN...
Run the query in RSRT and note the following.
OLAP time ( will be minimal since there is no processing as per your post)
DBTIME ( As Anil Kumar Sharma has mentioned )
To reduce DBTIME ..
1. Cache the query ( run it in the background using Reporting agent and turn on cache using RSRCACHE .. but then caching is not very effective for very large queries)
2. Create aggregates ( this way you can save time on accessing only query specific data )
3. Create secondary indexes on the relevant dimensions ( Again as previously mentioned )
4. See if you can keep most dimensions in the query in free characteristics and then allow the user to drill down ( If the report format is not chiselled in stone )
Arun
Assign points if found useful
You must be a registered user to add a comment. If you've already registered, sign in. Otherwise, register and sign in.
We're having the same problem. It's not the database or OLAP time, it's the front-end, when the data's been transferred to Excel. If Bex Properties > 'Adjust formatting after data refresh' is turned off it runs instantly but the query's left unformatted (including decimals and %'s), and then running a macro to format runs much quicker than Bex to look after the formatting, but doing it via a macro isn't a nice solution because it needs to handle the column filtering users may do on the query.
Any other solutions?
Regards,
Julie
maybe below links on query optimization could help..
/people/prakash.darji/blog/2006/01/26/query-optimization
/people/vikash.agrawal/blog/2006/04/17/query-performance-150-is-aggregates-the-way-out-for-me
/people/vikash.agrawal/blog/2006/04/26/in-what-147sequence148-olap-processor-look-for-query-relevant-data
cheers,
Vishvesh
You must be a registered user to add a comment. If you've already registered, sign in. Otherwise, register and sign in.
User | Count |
---|---|
85 | |
10 | |
10 | |
9 | |
7 | |
7 | |
6 | |
5 | |
4 | |
4 |
You must be a registered user to add a comment. If you've already registered, sign in. Otherwise, register and sign in.