Comparing column data in 2 tables with the same schema containing 1.5 billion records

Question

Post reply

Comparing column data in 2 tables with the same schema containing 1.5 billion records

JayK

SSCrazy

Points: 2679
More actions
April 14, 2013 at 5:07 pm

#271059

I have had the task of recreating a table in our data warehouse which contains 1.5 billions rows to make use of partioning on the clustered index by date. To keep the tables in sync over the past few weeks I have had to run import jobs on both tables.
Before moving across to the new partitioned table (by renaming it to bring it into production) I need to ensure that the EffectiveEndDate column in both tables is identical for every row (which has a PK).
So my question is how best is it to compare 2 tables with 1.5 billion records to ensure the data in both is the same?
At this stage 3rd party tools are not an option and I am using SQL Server 2012 SP1 Enterprise Edition.
Any help greatly appreciated!!

Viewing 5 posts - 1 through 4 (of 4 total)

You must be logged in to reply to this topic. Login to reply

Perry Whittle SSC Guru Points: 234013 More actions · Answer 1

are you saying you just want to check the data is identical for each row across all columns?

-----------------------------------------------------------------------------------------------------------

"Ya can't make an omelette without breaking just a few eggs" 😉

JayK SSCrazy Points: 2679 More actions · Answer 2

Hi Perry,

Thanks for getting back - in reality I just need to check that data in a single column is the same for each matching PK in the 2 tables.

So each table as an ID and an EffectiveEndDate column - each table will have the same number of rows - approx 1.5 billions rows. They should be a copy of each other but I need to ensure for each row has the same value for EffectiveEndDate in both tables,

Thanks again for your reply,

JK

ChrisM@Work SSC Guru Points: 186127 More actions · Answer 3

JayK (4/14/2013)
I have had the task of recreating a table in our data warehouse which contains 1.5 billions rows to make use of partioning on the clustered index by date. To keep the tables in sync over the past few weeks I have had to run import jobs on both tables.
Before moving across to the new partitioned table (by renaming it to bring it into production) I need to ensure that the EffectiveEndDate column in both tables is identical for every row (which has a PK).
So my question is how best is it to compare 2 tables with 1.5 billion records to ensure the data in both is the same?
At this stage 3rd party tools are not an option and I am using SQL Server 2012 SP1 Enterprise Edition.
Any help greatly appreciated!!

Start simple. Create a unique index on each table with columns ID and EffectiveEndDate.

Then run SELECT

t2.ID, t1.ID,

t2.EffectiveEndDate, t1.EffectiveEndDate

FROM Table1 t1

INNER JOIN Table2 t2

ON t2.ID = t1.ID

AND t2.EffectiveEndDate <> t1.EffectiveEndDate

^{“Write the query the simplest way. If through testing it becomes clear that the performance is inadequate, consider alternative query forms.” - Gail Shaw}

For fast, accurate and documented assistance in answering your questions, please read this article.
Understanding and using APPLY, (I) and (II) Paul White
Hidden RBAR: Triangular Joins / The "Numbers" or "Tally" Table: What it is and how it replaces a loop Jeff Moden

Perry Whittle SSC Guru Points: 234013 More actions · Answer 4

JayK (4/15/2013)
Hi Perry,
Thanks for getting back - in reality I just need to check that data in a single column is the same for each matching PK in the 2 tables.
So each table as an ID and an EffectiveEndDate column - each table will have the same number of rows - approx 1.5 billions rows. They should be a copy of each other but I need to ensure for each row has the same value for EffectiveEndDate in both tables,
Thanks again for your reply,
JK

use a select query on the columns required from each of the tables and apply the except operator, any differences in the column data will be returned, like so

select somecolumn, anothercolumn from thetablea

except

select somecolumn, anothercolumn from thetableb

-----------------------------------------------------------------------------------------------------------

"Ya can't make an omelette without breaking just a few eggs" 😉