First, query the data segment with the capacity greater than 1G with statements.
1.select segment_name, sum (bytes)/1024/1024 from DBA _ segments group by segment _ name has sum (bytes)/1024 > 1000;
The following results were obtained:
SYS _ lob 0000 13609 1c 00003 $ $ 255332m
SYS _ lob 0000 13644 1c 00004 $ $ 7 170m
SYS_C0082042 1305M
SYS_C0080433 1340M
2. According to the lob segment, query which table the LOB segment belongs to.
select table_name,segment _ name from DBA _ lobs where segment _ name in(select segment _ name from DBA _ segments group by segment _ name having sum(bytes)/ 1024/ 1024 & gt; 1000; );
After investigation, it is known that XXX table occupies a lot of capacity.
3. Delete the duplicate data rows in the XXX table, but after deleting these duplicate data, the disk space will not be released.
4. Then release the space occupied by lob type data.
Alter table adu_* moves the tablespace BFPICK lob (content) and stores it as (tablespace BFPICK).
5. After that, the space occupied by duplicate data is released, and then the table is re-indexed.
Change the index * * * * to regenerate;