 | |
Purging data without disturbing the production.
comp.databases.ibm-db2 comp.databases.ibm-db2
 |
| | Thread Tools | Display Modes |
#1
 
| |||
| |||
|
Purging data without disturbing the production. -
04-08-2011
, 08:56 AM
#2
| |||
| |||
|
|
Hi, What is the best procedure to purge data from production db2 databases without interuption of service and also with minimal or no risk?. I need to make this as a weekly/monthly schedule. The data growth may vary from 2 million to 5 million for 15 days. Please help me out. |
We're definitely moving to date-range partitioning for our data
archiving...
In 30 seconds of discussion:
You can create a partitioned-table...
Each Partition is setup with its own date-range...Part1
January...Part2 February...or Part1 QTR1, Part2 QTR2, etc.
You can at some point 'detach' a prior unsed partition (like January
or QTR2) and call it something ... 'frignitz' if you want.
Then either 'db2 drop table frignitz' or export it to another
database, etc.
You can create new partitions Part3 March...QTR3 for new data.
There should be no application-code changes.
Awesome.
-B
#3
| |||
| |||
|
|
On Apr 8, 9:56 am, kavin<kavinilammur... (AT) gmail (DOT) com> wrote: Hi, What is the best procedure to purge data from production db2 databases without interuption of service and also with minimal or no risk?. I need to make this as a weekly/monthly schedule. The data growth may vary from 2 million to 5 million for 15 days. Please help me out. Hello Kavin - We're definitely moving to date-range partitioning for our data archiving... In 30 seconds of discussion: You can create a partitioned-table... Each Partition is setup with its own date-range...Part1 January...Part2 February...or Part1 QTR1, Part2 QTR2, etc. You can at some point 'detach' a prior unsed partition (like January or QTR2) and call it something ... 'frignitz' if you want. Then either 'db2 drop table frignitz' or export it to another database, etc. You can create new partitions Part3 March...QTR3 for new data. There should be no application-code changes. Awesome. -B You can also take a look at IBM's Data Archiving/Data Growth solutions: |
http://www-01.ibm.com/software/data/...e-data-growth/
Larry Edelstein
#4
| |||
| |||
|
|
On Apr 8, 9:56 am, kavin<kavinilammur... (AT) gmail (DOT) com> wrote: Hi, What is the best procedure to purge data from production db2 databases without interuption of service and also with minimal or no risk?. I need to make this as a weekly/monthly schedule. The data growth may vary from 2 million to 5 million for 15 days. Please help me out. Hello Kavin - We're definitely moving to date-range partitioning for our data archiving... In 30 seconds of discussion: You can create a partitioned-table... Each Partition is setup with its own date-range...Part1 January...Part2 February...or Part1 QTR1, Part2 QTR2, etc. You can at some point 'detach' a prior unsed partition (like January or QTR2) and call it something ... 'frignitz' if you want. Then either 'db2 drop table frignitz' or export it to another database, etc. You can create new partitions Part3 March...QTR3 for new data. There should be no application-code changes. Awesome. -B You can also take a look at IBM's Data Archiving/Data Growth solutions: |
http://www-01.ibm.com/software/data/...e-data-growth/
Larry Edelstein
|
« Previous Thread
|
Next Thread »
| Thread Tools | |
| Display Modes | |
Linear Mode
| |
Powered by vBulletin Version 3.5.3
Copyright ©2000 - 2012, Jelsoft Enterprises Ltd.
Copyright ©2000 - 2012, Jelsoft Enterprises Ltd.