TimesTen SQL with group by returning multiple rows

Question

I have a Active-Standby TimesTen nodes.
Using group by with or without having clause:
Whenever I do a group by query on table1 table with or without having clause, SQL returns multiple rows. This looks very strange to me. Each time it gives different count
Command> select count(*) from table1 group by pname having pname='pool';
< 390400 >
1 row found.
Command> select count(*) from table1 group by pname having pname='pool';
< 390608 >
< 32639 >
2 rows found.
Command> select count(*) from table1 group by pname having pname='pool';
< 2394 >
< 351057 >
2 rows found.
Command> select count(*) from table1 group by pname having pname='pool';
< 305732 >
1 row found.
Command> select count(*) from table1 group by pname having pname='pool';
< 420783 >
1 row found.
Command> select count(*),pool_name from root.rms_address_pools group by pool_name order by pool_name;
< *435473, pool* >
< *32313, pool* >
< 453, smvG3 >
< *28980, pool* >
< 3786, smvG4 >
< *26025, pool* >
< 236120, smvG6 >
< 131455, smcG3 >
< *65150, pool* >
< 23, snt1G1 >
< 510, snt2G1 >
< 510, snt2G2 >
Using where clause:
Command> select count(*) from table1 where pname='pool';
< *442354* >
1 row found.
Command> select count(*) from table1 where pname='pool';
< 442354 >
1 row found.
Table description:
Command> desc table1;
Table table1:
Columns:
*IP_ADDRESS                      BIGINT NOT NULL
PNAME CHAR (32) NOT NULL
SITEID TINYINT NOT NULL
1 table found.
ttVersion:
bash-3.00# ./ttVersion
TimesTen Release *7.0.3.1.0 (64 bit Solaris)* (tt70:17001) 2007-10-30T22:17:07Z
Instance admin: root
Instance home directory: /TimesTen/tt70
Daemon home directory: /var/TimesTen/tt70
bash-3.00#
Could any one suggest what is wrong with my SQL? or is it a bug with TimesTen?
Many thanks in advance.
Br,
Brij

Answer 1

Hi Gena,
When i execute the query with where clause, it gives me the output with more than one pool:
Command> select pname, count (*) from table1 where pname='pool' group by pname ;
< smcG3 , 18836 >
< pool , 423527 >
2 rows found.
Command> select pname, count (*) from table1 where pname='pool' group by pname ;
< intG302 , 17202 >
< pool , 425159 >
2 rows found.
While if give use the having clause it gives me multiple rows for one pool only ( sometimes) :
select pname, count (*) from table1 group by pname having pname='pool';
< pool , 32686 >
< pool , 420445 >
2 rows found.
select pname, count (*) from table1 group by pname having pname='pool';
< pool , 393574 >
< pool , 5838 >
< pool , 110943 >
3 rows found.
Command> select pname, count (*) from table1 group by pname having pname='pool';
< pool , 414590 >
< pool , 8395 >
2 rows found.
Please suggest what can be done in this case, need i open a case with Oracle for this.
Regards, Brij