The search conditions vary depending on the type of object that you
want information about:
With no or old RUNSTATS information
This option shows objects that lack or have outdated RUNSTATS information.
Tip: Consider running the RUNSTATS utility on objects that match
this query.
With clustering level problems
This option shows objects that have more than 10 percent of their rows
in a far offset position or it shows objects that are defined as clustering
but have a clustering ratio less than 95 percent.
Tip: Consider reorganizing tables or redesigning indexes, tables,
or your applications. When redesigning objects and programs to avoid clustering
level problems, consider the following factors:
-
Insert, update, and delete patterns
-
Insert, update, and delete frequencies
-
Free space and reorganize frequencies
-
Clustering sequences
With index levels greater than 'n'
This option shows indexes that have a level greater than the value
that you specify in the Number of index levels field. Index
levels that exceed 2 or 3 can reduce the performance of your applications.
Tip: Consider reorganizing the indexes or redesigning the indexes
and tables. When reorganizing these indexes or redesigning these indexes
and tables, consider the following factors:
-
Key lengths
-
Free space
-
Insert, update, and delete patterns
-
Insert, update, and delete frequencies
Indexes with large leaf pages distance
This option shows indexes that have large leaf page distances. Generally,
leaf page distances greater than 200 might indicate a disorganized index.
This rule is a general indicator, and in some cases large leaf distances
are acceptable.
Which are type 1
This option shows indexes that are type 1.
Tip: Consider converting the index to a
type 2 index.
With long RID chains
This option shows indexes that have long record identifier (RID) chains,
which reduce performance during delete operations and during updates of
the indexed data. Long RID chains result from a low number of distinct
key values in relation to the total number of rows in the table.
Tip: Consider redesigning the index so
that the key values are more likely to be distinct and so that the index
is a type 2 index.
Not used by any plan or package with static SQL
This option shows indexes that are not used by any packages or plans.
Tip: If these indexes are not used in QMF or any other dynamic
SQL, then you should consider dropping them.
Without STOSPACE information
This option shows objects that do not have STOSPACE information.
Tip: To collect STOSPACE information, run the STOSPACE utility.
Consider running the STOSPACE utility periodically.
Exceeding allocated primary quantity
This option shows spaces that exceed the primary allocation.
Tip: If your query finds spaces that exceed primary quality,
consider extending the primary quality.
With more than 'n' percent relocated rows
This option shows spaces in which more than 'n' percent of the rows
do not reside in their original pages. These rows have been relocated.
Tip: Consider reorganizing the spaces or redesigning the program
that update their rows.
With more than 'n' percent dropped space
This option shows spaces in which more than 'n' percent of the space
contains dropped tables. This dropped space cannot be reused.
Tip: Consider reorganizing the spaces or using segmented table
spaces for the tables.
With table space locking
This option shows spaces that have the locking rule of 'S'. When
the locking rule equals 'S', DB2 uses table space locking when accessing
tables in the space.
Tip: You should use table space locking for only read-only tables
or tables that are accessed by one user (or batch job) at a time. Otherwise
consider changing the locking rule to 'A' by altering the table's lock
size.
Which are non-segmented and contain more than one
table
This option shows non-segmented spaces that contain more than one table.
Tip: In most cases, such spaces should contain only a single
table. Unless your design calls for more than one table in a non-segmented
space, consider moving the tables to separate spaces.