插入时，Sql select在表格上很慢

Question

I have a background job (C# console application) which inserts continuously data in product_titleParts table. Simply this job selects top 100 products from products table, splits title into part and inserts into product_titleParts table.

This table has a indexed column called "TitlePart"

On the other hand, when I try to select from this table SQL query takes forever. If I stop console app after a while query takes 0 seconds. As soon as I start console app again select query unresponsive again.

Selecting from a table while another job is inserting would cause any slowness? I am using nolock on the select but did not help.

Any Idea?

My Code on Console App:

if not exists(select 1 from product_titleParts where productid = @productid
and UserId = @userid and titlePart = @titlePart)
begin
insert into product_titleParts (userid, productid, titlePart)
VALUES (@userid, @productid, @titlePart)
end

My Select Code:

select productid from product_titleParts 
inner join products
on products.productid = product_titleParts.productid
where titlePart = @titlePart

Answer 1

I had simmilar problems with inserting "mass data" while selecting on the other hand (DB/2)

what was my solution:

inserts are expensive, so i tied them to a block like:

 insert into ...
 values ((a,b,c),(x,y,z), ... )

that locks the table only once for eg. 100 inserts and indices are only rebuild once

And there is an other point: when you are inserting (updataing) many rows, try to lock the table (depends on DBMS) in exclusive mode. You prevent whats called "lock escalation"

in DB2 you would use

   lock table xyz in exclusive mode
   update/insert many rows
   unlock table xyz