Re: Performance Problem with postgresql 9.03, 8GB RAM,Quadcore Processor Server--Need help!!!!!!!

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

 



Thanks Alban & Gregg.


i will describe little more about that table

                               Table "_100410.stk_source"
        Column         |   Type    |                      Modifiers                     
-----------------------+-----------+-----------------------------------------------------
 source_id             | integer   | not null default nextval('source_id_seq'::regclass)
 stock_id              | integer   |
 source_detail         | integer[] |
 transaction_reference | integer   |
 is_user_set           | boolean   | default false


We store transaction_type and transaction_id in source_detail column which is an interger array for each transactions

We use various functions to get the info based on transaction type

For eg:

In function to get the batch details we have used as

FOR batch_id_rec in select distinct(batch_id) from order_status_batches osb join batch_status_stock bss on osb.status_id=bss.batch_status_id where stock_id in (select source_detail[2] from stk_source where stock_id IN (SELECT std_i.stock_id                                                                                                                                     
        FROM order_details_shipments ods                                                                                                                                                       
        JOIN shipment_pack_stock sps ON sps.pack_id=ods.pack_id AND ods.order_id=sps.order_id AND ods.item_id=sps.item_id
        JOIN stock_transaction_detail_106 std ON std.transaction_id=sps.transaction_id
        JOIN stock_transaction_detail_106 std_i ON std.stock_id = std_i.stock_id AND std_i.transaction_type = 'i'::bpchar
        WHERE shipment_item_id=$1 ) and source_detail[1]=3) LOOP

...............................

................................

......................................

Similarly we have used in php pages and views

SELECT abd.bill_no as bill_no,to_char(abd.bill_date,'dd/mm/yyyy') AS date,mp.product_desc as product_desc,std.quantity,std.area,rip.price AS rate,
FROM acc_bill_items_106 abi
    JOIN acc_bill_details_106_table abd ON abd.bill_id=abi.bill_id AND abd.bill_status='act'
    JOIN stk_source_table ss ON ss.source_detail[2]=abi.item_id and ss.source_detail[1]=1
    JOIN stock_transaction_detail_106_table std ON std.stock_id=ss.stock_id
    JOIN stock_details_106_table sd106 ON sd106.stock_id=std.stock_id
    JOIN master_product_106_table mp ON mp.product_id= sd106.product_id
    JOIN receipt_item_price_106_table rip ON rip.receipt_item_id=abi.item_id
    WHERE abi.bill_id=$bill_id AND std.transaction_type='o'  ;

So where ever we have JOIN or used in functions the performance is very low some times query returns results takes more than 45 mints.

Normally if we fetch Select * from some_table..........it returns very fast because it has less records.

But when i put Select * from stk_source or to find the actual_cost

EXPLAIN ANALYZE SELECT * FROM stk_source;

i couln't able to retrieve the planner details waited for more than 50 to 60 mints

so question is in spite of having good server with high configuration and also changed the postgresql configuration settings then why the system is crawling?


What are the other parameters have to look out or what are the other config settings to be change to have the best performance??

Kindly help to sort out this problem......


Thanks in advance..................!!!!!!

Regards
Hashim




 


On Fri, Oct 28, 2011 at 5:07 PM, Alban Hertroys <haramrae@xxxxxxxxx> wrote:
On 28 October 2011 09:02, Mohamed Hashim <nmdhashim@xxxxxxxxx> wrote:
> EXPLAIN select * from stk_source ;
>                                      QUERY
> PLAN
> -------------------------------------------------------------------------------------
>  Result  (cost=0.00..6575755.39 rows=163132513 width=42)
>    ->  Append  (cost=0.00..6575755.39 rows=163132513 width=42)
>          ->  Seq Scan on stk_source  (cost=0.00..42.40 rows=1080 width=45)
>          ->  Seq Scan on stk_source  (cost=0.00..20928.37 rows=519179
> width=42)
>          ->  Seq Scan on stk_source  (cost=0.00..85125.82 rows=2111794
> width=42)
>          ->  Seq Scan on stk_source  (cost=0.00..6469658.80 rows=160500460
> width=42)

That plan gives you the best possible performance given your query.
Your example probably doesn't fit the problem you're investigating.

--
If you can't see the forest for the trees,
Cut the trees and you'll see there is no forest.



--
Regards
Mohamed Hashim.N
Mobile:09894587678

[Postgresql General]     [Postgresql PHP]     [PHP Users]     [PHP Home]     [PHP on Windows]     [Kernel Newbies]     [PHP Classes]     [PHP Books]     [PHP Databases]     [Yosemite]

  Powered by Linux