Output of the estimate compression operation (SQL administration API)
After you run the command for estimating compression ratios, the database server displays information that shows the estimate of the compression ratio that can be achieved, along with the currently achieved compression ratio (if it exists).
Column | Information Displayed |
---|---|
est |
This is the estimate of the compression ratio that can be achieved with a new compression dictionary. The estimate is a percentage of space saved compared to no compression. |
curr |
This is the estimate of the currently achieved
compression ratio. This estimate is a percentage of space saved compared
to no compression. 0.0% will always appear for non-compressed
fragments or tables. |
change |
This is the estimate of the percentage point
gain (or possibly loss, although that should be rare) in the compression
ratio that you could achieve by switching to a new compression dictionary.
This is just the difference between est and curr . If the table or fragment is not compressed, you can create a compression dictionary with the compress parameter. If the fragment is compressed, you must perform an uncompress or uncompress_offline operation, before you can compress. |
partnum |
This is the partition number of the fragment. |
coloff |
This value defines whether the estimate is for
in-row data or simple large objects in the dbspace, as follows:
|
table |
This is the full name of the table to which
the fragment belongs, in format database:owner.tablename If you are estimating compression benefits for an index, the full name of the index appears in this column. |
Example
The following output shows that
a .4 percent increase in saved space can occur if you recompress the
first fragment. A 75.7 percent increase can occur if you compress
the second fragment, which is not compressed. The value -1 in the coloff
column indicates
that in-row data is compressed.
est curr change partnum table
----- ----- ------ ------------- -----------------------------------
75.7% 75.3% +0.4 0x00200003 insurance:bwilson.disability
75.7% 0.0% +75.7 0x00300002 insurance:pchang.workcomp
est curr change partnum coloff table
----- ----- ------ ------------- -----------------------------
75.7% 75.3% +0.4 0x00200003 -1 insurance:bwilson.auto
75.7% 0.0% +75.7 0x00300002 -1 insurance:pchang.home
The following output shows compression estimates for in-row data (in the first row) and simple large objects at offsets 4 and 60 (in the second and third rows):
est curr change partnum coloff table
----- ----- ------ ------------- --------------------
75.4% 71.5% +3.9 0x00200002 -1 test:mah.table1
5.0% 75.0% +0.0 0x00200002 4 test:mah.table1
75.0% 75.0% +0.0 0x00200002 60 test:mah.table1
Output from compression estimates for tables and fragments look the same, except that the output for a table always shows all fragments in the table, while the output for a fragment only shows information for the specified fragments.