I am trying to write in a file stored in c:\ drive named vin1.txt and getting this error .Please suggest!
> ERROR at line 1: ORA-29280: invalid
> directory path ORA-06512: at
> "SYS.UTL_FILE", line 18 ORA-06512: at
> "SYS.UTL_FILE", line 424 ORA-06512: at
> "SCOTT.SAL_STATUS", line 12 ORA-06512:
> at line 1
HERE is the code
create or replace procedure sal_status
(
p_file_dir IN varchar2,
p_filename IN varchar2)
IS
v_filehandle utl_file.file_type;
cursor emp Is
select * from employees
order by department_id;
v_dep_no departments.department_id%TYPE;
begin
v_filehandle :=utl_file.fopen(p_file_dir,p_filename,'w');--Opening a file
utl_file.putf(v_filehandle,'SALARY REPORT :GENERATED ON %s\n',SYSDATE);
utl_file.new_line(v_filehandle);
for v_emp_rec IN emp LOOP
v_dep_no :=v_emp_rec.department_id;
utl_file.putf(v_filehandle,'employee %s earns:s\n',v_emp_rec.last_name,v_emp_rec.salary);
end loop;
utl_file.put_line(v_filehandle,'***END OF REPORT***');
UTL_FILE.fclose(v_filehandle);
end sal_status;
execute sal_status('C:\','vin1.txt');--Executing
Don't forget also that the path for the file is on the actual oracle server machine and not any local development machine that might be calling your stored procedure. This is probably very obvious but something that should be remembered.
For utl_file.open(location,filename,mode) , we need to give directory name for location but not path. For Example:DATA_FILE_DIR , this is the directory name and check out the directory path for that particular directory name.
You need to have your DBA modify the init.ora file, adding the directory you want to access to the 'utl_file_dir' parameter. Your database instance will then need to be stopped and restarted because init.ora is only read when the database is brought up.
You can view (but not change) this parameter by running the following query:
Share and enjoy.
The directory name seems to be case sensitive. I faced the same issue but when I provided the directory name in upper case it worked.
Since Oracle 9i there are two ways or declaring a directory for use with UTL_FILE.
The older way is to set the INIT.ORA parameter UTL_FILE_DIR. We have to restart the database for a change to take affect. The value can like any other PATH variable; it accepts wildcards. Using this approach means passing the directory path...
The alternative approach is to declare a directory object.
Directory objects require the exact file path, and don't accept wildcards. In this approach we pass the directory object name...
The UTL_FILE_DIR is deprecated because it is inherently insecure - all users have access to all the OS directories specified in the path, whereas read and write privileges can de granted discretely to individual users. Also, with Directory objects we can be add, remove or change directories without bouncing the database.
In either case, the
oracle
OS user must have read and/or write privileges on the OS directory. In case it isn't obvious, this means the directory must be visible from the database server. So we cannot use either approach to expose a directory on our local PC to a process running on a remote database server. Files must be uploaded to the database server, or a shared network drive.If the
oracle
OS user does not have the appropriate privileges on the OS directory, or if the path specified in the database does not match to an actual path, the program will hurl this exception:The OERR text for this error is pretty clear:
You need to register the directory with Oracle. fopen takes the name of a directory object, not the path. For example:
(you may need to login as SYS to execute these)
Then, you can refer to it in the call to fopen: