catch.saravana
Member
Hello Everyone,
I have a large table which is of size 300GB - like an audit table. We wanted to do a D&L for this table separately and not as part of our migration. I assume we will not be able to do a D&L in parallel with users writing data to this table.
What we are trying to get here is;
1. Load the schema which will include this large table
2. D&L for all tables other than the large one
3. Index rebuild for all tables
4. Start the DB
5. Users will connect to the application and start using it
6. Data will be written to the large table as well which is empty right now
7. In parallel I have to load the 300 GB data to that large table
Will table partition help? Or is there a better way to achieve this kind of a scenario?
If we don't have a better way then we are planning to the follow the below plan;
1. We can rename the large table (Let's say tbl will be renamed as artbl)
2. We will now have artbl which is 300 GB and tbl which is size 0
3. Users will be writing data to tbl
4. Users will be reading data from both artbl and tbl to fetch the right data
Note - to get this working we will need to modify 45 programs, so rather going with the above approach we wanted to see if we have a better solution.
I have a large table which is of size 300GB - like an audit table. We wanted to do a D&L for this table separately and not as part of our migration. I assume we will not be able to do a D&L in parallel with users writing data to this table.
What we are trying to get here is;
1. Load the schema which will include this large table
2. D&L for all tables other than the large one
3. Index rebuild for all tables
4. Start the DB
5. Users will connect to the application and start using it
6. Data will be written to the large table as well which is empty right now
7. In parallel I have to load the 300 GB data to that large table
Will table partition help? Or is there a better way to achieve this kind of a scenario?
If we don't have a better way then we are planning to the follow the below plan;
1. We can rename the large table (Let's say tbl will be renamed as artbl)
2. We will now have artbl which is 300 GB and tbl which is size 0
3. Users will be writing data to tbl
4. Users will be reading data from both artbl and tbl to fetch the right data
Note - to get this working we will need to modify 45 programs, so rather going with the above approach we wanted to see if we have a better solution.