All,
I am working on a new requirement to build a very large table, 1 million records per day, 180 day retention.
To access these records quickly and efficiently, it was mentioned that the table should be partitioned by date. At the start of each day, a stored procedure would be run that would first drop the partition (and index) that has now become 181 days old. It would then create a new partition and index for the current day.
I thought I tried doing this a while back and could not get it to work. I was using OEM to create the partitioned table, and I don't think it would take a date as the partitioned field.
Anyone else have any luck with this? Anyone see any problems with this sort of design?
Terry
**************************
* General Disclaimer - Please read *
**************************
Please make sure your post is in the CORRECT forum, has a descriptive title, gives as much detail to the problem as possible, and has examples of expected results. This will enable me and others to help you faster...
I am working on a new requirement to build a very large table, 1 million records per day, 180 day retention.
To access these records quickly and efficiently, it was mentioned that the table should be partitioned by date. At the start of each day, a stored procedure would be run that would first drop the partition (and index) that has now become 181 days old. It would then create a new partition and index for the current day.
I thought I tried doing this a while back and could not get it to work. I was using OEM to create the partitioned table, and I don't think it would take a date as the partitioned field.
Anyone else have any luck with this? Anyone see any problems with this sort of design?
Terry
**************************
* General Disclaimer - Please read *
**************************
Please make sure your post is in the CORRECT forum, has a descriptive title, gives as much detail to the problem as possible, and has examples of expected results. This will enable me and others to help you faster...