Your high value for each partition can just be the beginning of every month. For
example:
CREATE TABLE ACCOUNTS
(
STATEMENT_DATE DATE NOT NULL,
ACCOUNT_NUMBER VARCHAR2(8)NOT NULL,
BILLING_CYCLE VARCHAR2(2)NOT NULL,
Resending, since my message was truncated...
Your high value for each partition can just be the beginning of every month. For
example:
CREATE TABLE ACCOUNTS
(
STATEMENT_DATE DATE NOT NULL,
ACCOUNT_NUMBER VARCHAR2(8)NOT NULL,
BILLING_CYCLE
Rick,
If you partition on the date field then you set the partition limit
with a date type limit.
PARTITION aadedup01 VALUES LESS THAN ('02-01-2003')
TABLESPACE aadedup01_data,
PARTITION aadedup02 VALUES LESS THAN ('03-01-2003')
TABLESPACE aadedup02_data,
PARTITION aadedup03 VALUES LESS THAN
From what I understood, he's asking how to put all January records in one partition,
all Feb in another etc.
So you end up with 12 partitions for as many years as you wish. I could not think of
any other solution then the one he does not want to.
rw
Your high value for each partition can just
Yes, this is what I was looking for. I don't care about the year, and don't
want to worry about adding new partitions for every new month that comes
along each year. This table will only needs to contain six months worth of
data. I will not be archiving it at all. I wanted to truncate the
Rick - Well, a simple way to do it might be to create 5 years of partitions,
based on date as described. Then map all the Jan partitions to the same
tablespace, Feb partitions to the same tablespace, etc. Instead of
truncating a partition, just drop it to reclaim the space. Crude, but might
meet
Alternatively you could create the table based on a time_key and map that to the time table. Then the partitions key could change meaning as the years go by.
i.e. time_key 1 = January of 2004
time_key 2 = Februrary of 2004
...
truncate the partitions as needed and reuse by changing the date in
And make sure all your indexes are local."M.Godlewski" [EMAIL PROTECTED] wrote:
Alternatively you could create the table based on a time_key and map that to the time table. Then the partitions key could change meaning as the years go by.
i.e. time_key 1 = January of 2004
time_key 2 = Februrary
create table aadedupekeys
(
file_id number,
rundate date,
pk number(10),
dk varchar2(128),
constraint aadedupekeys_pk primary key (file_id,rundate)
)
;
Is there anyway to partition a table (above) in months e.g.
('January','Februray'...). I want to use the rundate and
Hi gurus,
I have just migrated from Oracle 8.1.6 to Oracle 8.1.7 (Enterprise Edition on
Windows 2000) but when I try to query my partitoned tabless I get the error
ORA-00604 'error occured at recursive SQL level 1'
ORA -00904 'Invalid column name'
I have run the scripts U0801060.sql (which calls
Hi
We had a different problem with our upgrade but the same error message -
there is a patch available from Oracle that is supposed to help(did n't for
us though)
Regards,
N.
-Original Message-
[EMAIL PROTECTED]
Sent: 01 April 2003 09:04
To: Multiple recipients of list ORACLE-L
Hi
: Using Table Partitions in
8.0.5
Thats everyone. I just
found that out from Joe. This really sucks
.. Save the money .. Loose the
features. I hate limping along like this
-Original Message-From: Yuval Arnon [mailto:[EMAIL PROTECTED]]Sent:
Monday, August 20
Hi folks;
I have tried to setup a partitioned
table in 8.0.5 and I get the message
create table
calc_sched_geo_factor_p*ERROR at line 1:ORA-00439: feature not
enabled: Partitioning
Is there something someplace I need to turn
on in order to use partitions ??
Thanks
Kevin
12:27
PMTo: Multiple recipients of list ORACLE-LSubject: Using
Table Partitions in 8.0.5
Hi folks;
I have tried to setup a partitioned
table in 8.0.5 and I get the message
create table
calc_sched_geo_factor_p*ERROR at line 1:ORA-00439: feature not
enabled: Partitioning
14 matches
Mail list logo
