| Oracle FAQ | Your Portal to the Oracle Knowledge Grid | |
Home -> Community -> Usenet -> c.d.o.server -> Contents of DBA_HISTOGRAMS according to the ANALYZE cmd used
Here is my table (all types are NUMBER):
SQL> desc sc.graphe
Name
SQL> break on column_name
SQL> select column_name,ENDPOINT_NUMBER,
2 ENDPOINT_VALUE , ENDPOINT_ACTUAL_VALUE eav from dba_histograms 3 where table_name='GRAPHE' and owner='SC' order by 1,2,3;
COLUMN_NAME ENDPOINT_NUMBER ENDPOINT_VALUE EAV
------------------------- --------------- -------------- ---
DISTANCE 10 0
11 15
12 40
15 50
16 54
[...]
73 8850
74 11853
75 38100
END_NODE 0 52
1 172
2 321
3 519
4 718
5 887
6 1123
7 1369
8 1725
9 2048
10 2284
[...]
74 4491
75 4554
JOB_ID 138 11856
282 11926
426 11930
586 14921
729 14924
903 14925
1050 14926
START_NODE 0 104
1 291
2 431
3 651
4 828
5 1050
[...]
73 4455
74 4488
75 4553
221 rows selected.
And here is what you get with a mere ANALYZE TABLE ESTIMATE STATISTICS:
SQL> select column_name ,ENDPOINT_NUMBER, ENDPOINT_VALUE , EAV
2 from dba_histograms [...]
COLUMN_NAME ENDPOINT_NUMBER ENDPOINT_VALUE ENDPOINT_ACTUAL_VALUE
---------------- --------------- -------------- ---------------------
DISTANCE 0 0
1 38100
END_NODE 0 52
1 4554
JOB_ID 0 11856
1 14926
START_NODE 0 104
1 4553
8 rows selected.
Hence my 2 questions:
1/ The doc. about the 3rd column says "Actual (not normalized) string
value of the endpoint for this bucket": what does it mean, and when
is it populated ??
2/ Why is DBA_HISTOGRAMS populated so differently whether you use a
command or the other and how should one interpret the values these rows contain ?
Thanks a lot.
Sp
Received on Mon Jan 31 2005 - 02:53:57 CST
![]() |
![]() |