Home » RDBMS Server » Performance Tuning » "db file scattered read" too high and Query going for full table scan-Why ?
"db file scattered read" too high and Query going for full table scan-Why ? [message #166024] Tue, 04 April 2006 02:07 Go to next message
tayalarun
Messages: 20
Registered: December 2005
Junior Member
Hi,

I had a big table of around 200mb and had a index on it.
In my query I am using the where clause which has to use the
index. I am neither using any not null condition
nor using any function on the index fields.

Still my query is not using the index.
It is going for full table scan.
Also the statspack report is showing the
"db file scattered read" too high.

Can any body help and suggest me why this is happenning.
Also tell me the possible solution for it.

Thanks

Arun Tayal
Re: "db file scattered read" too high and Query going for full table scan-Why ? [message #166039 is a reply to message #166024] Tue, 04 April 2006 03:44 Go to previous messageGo to next message
madhusunkara
Messages: 59
Registered: March 2006
Location: hyderabad,india
Member
do you have uptodate stats for table and indexes

Regards
Madhu
Re: "db file scattered read" too high and Query going for full table scan-Why ? [message #166040 is a reply to message #166039] Tue, 04 April 2006 03:49 Go to previous messageGo to next message
tayalarun
Messages: 20
Registered: December 2005
Junior Member
Hi Madhu

Yes I had done analyze ... compute statistics on both table and index.



Regards

Arun Tayal


Re: "db file scattered read" too high and Query going for full table scan-Why ? [message #166041 is a reply to message #166040] Tue, 04 April 2006 03:56 Go to previous messageGo to next message
madhusunkara
Messages: 59
Registered: March 2006
Location: hyderabad,india
Member
what is the slectivity for the index, may be fts is good in you case

Thanks & Regards
Madhu

[Updated on: Tue, 04 April 2006 03:59]

Report message to a moderator

Re: "db file scattered read" too high and Query going for full table scan-Why ? [message #166151 is a reply to message #166024] Tue, 04 April 2006 18:11 Go to previous message
RaeMarvin
Messages: 7
Registered: April 2006
Junior Member
To help could you post the stats details below.

Index: Index_name, blevel, leaf_blocks, distinct_keys, clustering_factor, avg_leaf_blocks_per_key, avg_datta_blocks_per_key.

Table: blocks, num_rows, freelists
column: column_name, density, num_nulls

The analyze command your using, the predicates at least from the query, version of Oracle as this has major impact on the plan and the explain plan.

From that information somone might be able to aid more.

But as stated don't be surprised if a FTS is the deemed as th best path.

Cheers
Rae
Previous Topic: gather_fixed_objects_stats
Next Topic: slow connection to oracle 8 server with oracle 9 client
Goto Forum:
  


Current Time: Wed Nov 27 08:28:17 CST 2024