Unanswered: Unable to extend error occurs on HP-UNIX machine though there is ample free space
Oracle 22.214.171.124 on HP-UNIX machine.
We have created table space
CREATE TABLESPACE TS_DATA
DATAFILE '/u03/oracle/oradata/TS_DATA/TS_DATA.dbf' SIZE 25600K AUTOEXTEND ON NEXT 1024K MAXSIZE UNLIMITED
EXTENT MANAGEMENT LOCAL AUTOALLOCATE
SEGMENT SPACE MANAGEMENT AUTO
Now when we are importing data the dbf file grows but it stops at around 1670MB and gives the following error.
IMP-00058: ORACLE error 1653 encountered
ORA-01653: unable to extend table TEST_USER.EMP by 1024 in tablespace TS_DATA
ORA-27059: skgfrsz: could not reduce file size
HP-UX Error: 27: File too large
Additional information: 2
IMP-00028: partial import of previous table rolled back: 830791 rows rolled back
About to enable constraints...
Import terminated successfully with warnings.
We have nearly 100GB OS disk space free allocated to this mount point.
This problem gets solved by adding one more data file to the same tablespace but we need to know why the first data file is not extending beyond a certain size.
This problem occured for 3 different databases.Database 1 it gave the error at 1670 MB,Database 2 gave error at 1900MB and Database 3 gave error at 2090MB.
We are not able to understand why it stops at a particular size and gives this error.
Pl. respond if you have come across the same situation and solved it.
We would like to have only one data file with autoextend.
The same setup works fine on Solaris and Windows machines but not on HP-UNIX machine.
We have used the same tablespace creation scripts for 2 other environments windows and solaris and it worked fine there.We were able to import large data.It was autoallocating properly.
But it is this OS which is giving this problem.
Can you give more details as to how set the maximum extents.
We had given to increase 1024k for each extend which ORACLE does.
Then if default number of extents was 4096 then
( 1024k * 4096 ) / 1024k=4096MB should have been allocated right?
But it has just allocated 2090M and stopped and this also differs for other 2 databases on the machine.
Pl. correct me if I am wrong.
I suspect we need some patch to be applied as I guess this is a bug with ORACLE.
We have oracle 126.96.36.199 ,can anyone tell me whether after upgrading to oracle 188.8.131.52 this problem will be rectified.
Excuse me for my late response, but I've been out of town on a business trip. I am not fluent in HP-UNIX, but the "lsfs -q <file system>" command shows on AIX the information requested under the "bf" property.