Bill, Its possible that developers forgot how it was created .. or they might have created it from a table which had the problem in the first place. The key point is that the problem exists and the resolution is noted in the metalink Note... For example: ORA92> create table x1 as select * from x1@remotedb; Table created. ORA92> desc x1 Name Null? Type ----------------------------------------------------------------- -------- -------------------------------------------- A NUMBER B TIMESTAMP(6) ORA92> select column_name, data_type, data_length, data_scale from user_tab_columns where table_name = 'X1'; COLUMN_NAME DATA_TYPE DATA_LENGTH DATA_SCALE -------------------- -------------------- ----------- ---------- A NUMBER 22 B TIMESTAMP(0) 11 0 ORA92> create table x2 as select * from x1; Table created. ORA92> desc x2 Name Null? Type ----------------------------------------------------------------- -------- -------------------------------------------- A NUMBER B TIMESTAMP(6) ORA92> select column_name, data_type, data_length, data_scale from user_tab_columns where table_name = 'X2'; COLUMN_NAME DATA_TYPE DATA_LENGTH DATA_SCALE -------------------- -------------------- ----------- ---------- A NUMBER 22 B TIMESTAMP(0) 11 0 Now I can say that X2 was not created via CTAS from remote db link :) Let us know if the fix in that note solved your issue ... I can see this bug to potentially cause issues for folks who are using exp/imp to upgrade dbs... Anurag On 3/2/07, William Wagman <wjwagman@xxxxxxxxxxx> wrote:
Anurag, I saw that note but I did not try it as I was assured by the developers that the table was not created using CTAS via a database link. I guess I should just go ahead and try it anyway. Thanks. Bill Wagman Univ. of California at Davis IET Campus Data Center wjwagman@xxxxxxxxxxx (530) 754-6208