|
View previous topic :: View next topic
|
| Author |
Message |
pshongal
New User
Joined: 14 Jun 2012
Posts: 96
Location: India
|
|
|
|
Hi,
We are inserting huge amount of data in a table. We are keeping the data in table for 1 month. Amount of data inserted is around 15MM records per day and length of record is about 2500 (variable).
My question is, whether dumping this data in a TAPE weekly and keeping only 1 week of data in table will reduce the cost? I want to know TAPE is cheaper or DB2 table?
Thanks.
PS |
|
| Back to top |
|
 |
Bill Woodger
Moderator Emeritus
Joined: 09 Mar 2011
Posts: 7309
Location: Inside the Matrix
|
|
|
|
The answer to that should be obvious to you.
The real question is, if the intention is to hold the data for one month, is it really likely that for the last three weeks that tape is an acceptable medium for data-enquiry. This seems unlikely in general, but obviously it depends entirely on the specifics of the data. |
|
| Back to top |
|
|
Nic Clouston
Global Moderator
Joined: 10 May 2007
Posts: 2455
Location: Hampshire, UK
|
|
|
|
| How do you propose to query the data on the tape- it is no longer a DB2 table but a sequential dataset? You would have to write programs to run in batch to handle queries against that data. |
|
| Back to top |
|
|
pshongal
New User
Joined: 14 Jun 2012
Posts: 96
Location: India
|
|
|
|
Hi Bill/Nic,
We do not query the data regularly unless when there is a issue and need to check the data, and this scenario is very rare in our environment. So we are least bothered on data query. |
|
| Back to top |
|
|
|
|
