Is like that, I now in the maintenance of a data warehouse, and then the library every day there are a lot of the delete operation, a lot of space high water level is high,
Think the solution is: MOVE tables,
Problem: MOVE table takes time every week, is there any way, for example, change the database parameters such as, completely can't get rid of high water level data,

CodePudding user response:

This kind of situation can only backup regularly, to build table, then import to solve it

CodePudding user response:

Analysis whether can insert into the history table, then truncate table,

CodePudding user response:

Don't know the specific circumstances of the building Lord, perhaps a partitioned table can solve your problem,

If they are inserting data according to time, writing of data every day is also a time (the time), could be designed as follows:

1, when writing new data every day, to create a new partition, partition name can use date to distinguish,
2, delete data, two months ago, the what day directly to the partition deleted, fast,

CodePudding user response:

The move is the most convenient method
In fact, I doubt HWM have much impact on your application can?

CodePudding user response:

The

reference 4 floor minsic78 reply:

move is the most convenient method
In fact, I doubt HWM have much impact on your application can?

More care about this problem, never understand why only grow high water level,

CodePudding user response:

reference 5 floor flyfeifei66 reply:

Quote: refer to 4th floor minsic78 response:

The move is the most convenient method
In fact, I doubt HWM have much impact on your application can?

More care about this problem, never understand why only grow high water level,

That's for sure, otherwise how could called high water level

CodePudding user response:

The key lies in the business logic, data flow of normal, out into, generally do not have problems with high water
If you delete the account for most of the data, consider using truncate + insert needs to retain data for