![]() | |
![]() |
| | Thread Tools | Display Modes |
#1
| |||
| |||
|
#2
| |||
| |||
|
|
Can anyone tell me why the cube partition processing fails when I use a view as the fact table, and partition it with a data slice. We have one large fact table, which is created through a view. This works fine, but when I try to break the cube up with a data slice based on month, I get the following error: The object structure is not valid (Invalid fact table alias); I'm not finding much info on this error. Any help would be great.. thanks |
#3
| |||
| |||
|
|
allie; i think that you'll need to give more information can you give us the SQL definition for the view? Thanks "allie" <aleslie (AT) sympatico (DOT) ca> wrote in message news:eNPRpBj1EHA.2608 (AT) TK2MSFTNGP10 (DOT) phx.gbl... Can anyone tell me why the cube partition processing fails when I use a view as the fact table, and partition it with a data slice. We have one large fact table, which is created through a view. This works fine, but when I try to break the cube up with a data slice based on month, I get the following error: The object structure is not valid (Invalid fact table alias); I'm not finding much info on this error. Any help would be great.. thanks |
#4
| |||
| |||
|
|
The view which I am using to define the fact table is similar to something as follows: CREATE VIEW VW_FACT_TABLE AS SELECT g.DATE_KEY, g.TIME_BAND_KEY, g.APPLICATION_KEY, g.CUSTOMER_KEY, g.REGION_KEY, g.AREA_CODE_KEY, SOURCE_KEY, g.TABLE1_KEY, g.TABLE1_STATUS_KEY, g.REASON_KEY, g.TABLE1_COUNT, g.TABLE1_DURATION, g.CREATE_DATE FROM F_TABLE1_DETAIL g JOIN D_APPLICATION a ON g.APPLICATION_KEY = a.APPLICATION_KEY WHERE TYPE_FLAG <> '1' UNION ALL SELECT m.DATE_KEY, m.TIME_BAND_KEY, m.APPLICATION_KEY, m.CUSTOMER_KEY, m.REGION_KEY, 0, m.SOURCE_KEY, m.TABLE2_KEY, 0, m.REASON_KEY, m.TABLE2_COUNT, m.TABLE2_DURATION, m.CREATE_DATE FROM F_TABLE2_DETAIL m JOIN D_APPLICATION a ON g.APPLICATION_KEY = a.APPLICATION_KEY WHERE a.TYPE_FLAG = '2' OR a.TYPE_FLAG = '4' I've tried it on another server, and it does exactly the same thing when I create a partition and select a date slice. Then process the cube always fails with the same error -> The object structure is not valid (Invalid fact table alias); Please let me know if you can see any reason this will not be possible to set up in partitons. thanks for the response. aaron_kempf (AT) hotmail (DOT) com> wrote in message news:uYuv1Uk1EHA.1144 (AT) TK2MSFTNGP09 (DOT) phx.gbl... allie; i think that you'll need to give more information can you give us the SQL definition for the view? Thanks "allie" <aleslie (AT) sympatico (DOT) ca> wrote in message news:eNPRpBj1EHA.2608 (AT) TK2MSFTNGP10 (DOT) phx.gbl... Can anyone tell me why the cube partition processing fails when I use a view as the fact table, and partition it with a data slice. We have one large fact table, which is created through a view. This works fine, but when I try to break the cube up with a data slice based on month, I get the following error: The object structure is not valid (Invalid fact table alias); I'm not finding much info on this error. Any help would be great.. thanks |
#5
| |||
| |||
|
|
well, what fact Table aliases are you using? "allie" <aleslie (AT) sympatico (DOT) ca> wrote in message news:OSpyV$o1EHA.1452 (AT) TK2MSFTNGP11 (DOT) phx.gbl... The view which I am using to define the fact table is similar to something as follows: CREATE VIEW VW_FACT_TABLE AS SELECT g.DATE_KEY, g.TIME_BAND_KEY, g.APPLICATION_KEY, g.CUSTOMER_KEY, g.REGION_KEY, g.AREA_CODE_KEY, SOURCE_KEY, g.TABLE1_KEY, g.TABLE1_STATUS_KEY, g.REASON_KEY, g.TABLE1_COUNT, g.TABLE1_DURATION, g.CREATE_DATE FROM F_TABLE1_DETAIL g JOIN D_APPLICATION a ON g.APPLICATION_KEY = a.APPLICATION_KEY WHERE TYPE_FLAG <> '1' UNION ALL SELECT m.DATE_KEY, m.TIME_BAND_KEY, m.APPLICATION_KEY, m.CUSTOMER_KEY, m.REGION_KEY, 0, m.SOURCE_KEY, m.TABLE2_KEY, 0, m.REASON_KEY, m.TABLE2_COUNT, m.TABLE2_DURATION, m.CREATE_DATE FROM F_TABLE2_DETAIL m JOIN D_APPLICATION a ON g.APPLICATION_KEY = a.APPLICATION_KEY WHERE a.TYPE_FLAG = '2' OR a.TYPE_FLAG = '4' I've tried it on another server, and it does exactly the same thing when I create a partition and select a date slice. Then process the cube always fails with the same error -> The object structure is not valid (Invalid fact table alias); Please let me know if you can see any reason this will not be possible to set up in partitons. thanks for the response. aaron_kempf (AT) hotmail (DOT) com> wrote in message news:uYuv1Uk1EHA.1144 (AT) TK2MSFTNGP09 (DOT) phx.gbl... allie; i think that you'll need to give more information can you give us the SQL definition for the view? Thanks "allie" <aleslie (AT) sympatico (DOT) ca> wrote in message news:eNPRpBj1EHA.2608 (AT) TK2MSFTNGP10 (DOT) phx.gbl... Can anyone tell me why the cube partition processing fails when I use a view as the fact table, and partition it with a data slice. We have one large fact table, which is created through a view. This works fine, but when I try to break the cube up with a data slice based on month, I get the following error: The object structure is not valid (Invalid fact table alias); I'm not finding much info on this error. Any help would be great.. thanks |
![]() |
| Thread Tools | |
| Display Modes | |
| |