On Fri, Feb 11, 2011 at 12:47 PM, Aaron Burnett wrote:
>
>
> Thank you all very much for your help.
>
> The suggestion from Osvaldo below was the best for my situation (not having
> any soret of xref table to join)...
>
>
It may work well for now, but if that foo_activity table has the potential
Thank you all very much for your help.
The suggestion from Osvaldo below was the best for my situation (not having any
soret of xref table to join)...
Best Regards,
Aaron
On 2/11/11 1:09 PM, "Osvaldo Kussama" wrote:
2011/2/11, Aaron Burnett :
>
> Hi,
>
> I'm just drawing a blank entirely
2011/2/11, Aaron Burnett :
>
> Hi,
>
> I'm just drawing a blank entirely today and would appreciate some help on
> this.
>
> The long and short; there are 12 distinct activities that need to be queried
> on a weekly basis:
>
> SELECT count(activity_id), activity_id
> FROM foo_activity
> WHERE creat
Assuming you have a table which lists all possible activities, with one
activity per row and no duplicates, you need to do a left outer join between
activities and your query result. That will generate a resultset that has
at least one row for every row in activities, with nulls in all the columns
On 02/11/2011 11:46 AM, Aaron Burnett wrote:
>
> Hi,
>
> I'm just drawing a blank entirely today and would appreciate some help on
> this.
>
> The long and short; there are 12 distinct activities that need to be queried
> on a weekly basis:
>
> SELECT count(activity_id), activity_id
> FROM fo
Not tested.
1. select count(t2.activity_id),
t1.activity_id
from (select distinct activity_id from foo_activity) as t1, -- assumes
all activities exist somewhere in table
left join foo_activity t2 on (t1.activity_id = t2.activity_id)
WHERE created >= '01/01/2011' and created < '01/08/20
Hi,
I'm just drawing a blank entirely today and would appreciate some help on
this.
The long and short; there are 12 distinct activities that need to be queried
on a weekly basis:
SELECT count(activity_id), activity_id
FROM foo_activity
WHERE created >= '01/01/2011' and created < '01/08/2011'
Even DB2 and Oracle will take hellishly long times to perform large
scale deletes
What I do for a database just under 300 gb in size is do deletes in
groups of 10,000
So your where clause might look some like
WHERE id NOT IN (SELECT id FROM unique_records fetch first 1 rows
only)
DB2
