October 12, 2012 at 3:33 am
Hi Folks,
We have a DWH database in which one of a table has more than 5 Crore of records. In that table, nearly 15 Laks records are getting loaded on daily basis.
We want to Archive [Not Deleting the real data's] some past days records from this table. For example, we need Archive July, August, September month data's.
Can i know which is the effective way to do this?
Thanks in Advance!
October 12, 2012 at 5:06 am
daily data of 15 lakhs has total of 5 crore means average of 30 -35 days of data only right.
do you have any keys / indexes in the table.
try moving old data to different DB and link it and use.
Regards
Durai Nagarajan
October 12, 2012 at 6:06 am
you can move old data to another table on same db. Create some job to achieve the same.
October 12, 2012 at 6:16 am
FYI: http://en.wikipedia.org/wiki/South_Asian_numbering_system
Edit: fixed link
October 12, 2012 at 6:23 am
homebrew01 (10/12/2012)
what is this?
Regards
Durai Nagarajan
October 12, 2012 at 6:57 am
sqlusers (10/12/2012)
Hi Folks,We have a DWH database in which one of a table has more than 5 Crore of records. In that table, nearly 15 Laks records are getting loaded on daily basis.
We want to Archive [Not Deleting the real data's] some past days records from this table. For example, we need Archive July, August, September month data's.
Can i know which is the effective way to do this?
Thanks in Advance!
Enterprise or Standard Edition?
--Jeff Moden
Change is inevitable... Change for the better is not.
October 12, 2012 at 7:02 am
Hi,
would this not be the perfect opportunity to partition your database. Select the criteria on which the partition would be created (ie: by date) and use an automated job to move the redundant data to another partition\storage device\table on a pretermined schedule.
At least that is what I would do......
Regards
Kev
October 12, 2012 at 7:12 am
durai nagarajan (10/12/2012)
homebrew01 (10/12/2012)
FYI: http://en.wikipedia.org/wiki/South_Asian_numbering_system%5B/quote%5Dwhat is this?
A link to a wiki page to help explain what a crore and a lak is. There are many of us that have no idea what these terms mean.
Let's make it easier for others to us this link:
October 12, 2012 at 7:12 am
kevaburg (10/12/2012)
Hi,would this not be the perfect opportunity to partition your database. Select the criteria on which the partition would be created (ie: by date) and use an automated job to move the redundant data to another partition\storage device\table on a pretermined schedule.
At least that is what I would do......
Regards
Kev
I was thinking the same thing when I asked which Edition of SQL Server they have. If Enterprise then Table Partion. If Standard Edition then Partioned View.
--Jeff Moden
Change is inevitable... Change for the better is not.
October 12, 2012 at 7:29 am
Lynn Pettis (10/12/2012)
durai nagarajan (10/12/2012)
homebrew01 (10/12/2012)
FYI: http://en.wikipedia.org/wiki/South_Asian_numbering_system%5B/quote%5Dwhat is this?
A link to a wiki page to help explain what a crore and a lak is. There are many of us that have no idea what these terms mean.
Let's make it easier for others to us this link:
oh sorry i thought he wrongly posted it..
Regards
Durai Nagarajan
October 12, 2012 at 11:02 am
Sorry for the delay response. 🙂
Yes, we have indexes and using standard edition. Seems different DB will not work out since permission issue.
Will try to automate as per the suggestions.
Thanks!
October 12, 2012 at 12:12 pm
sqlusers (10/12/2012)
Sorry for the delay response. 🙂Seems different DB will not work out since permission issue.
Thanks!
Then archiving probably isn't going to do anything for you. You just need to index correctly and then write effecient queries against it. To make index maintenance easier, consider using a Partitioned View (since you have standard edition).
--Jeff Moden
Change is inevitable... Change for the better is not.
Viewing 12 posts - 1 through 11 (of 11 total)
You must be logged in to reply to this topic. Login to reply