Modifying SQL Query - sql

Question

Previously i wanted to find
1- Batch id's
2- Upload time
3- Count of requests infront of each batch
i wrote the following query and it worked perfectly
SELECT REQUEST_BATCH_ID,To_CHAR(B.UPDATE_STAMP , 'DD/MM/YYYY HH24'),COUNT(*)
FROM bpdata.bulk_prov_detail B
WHERE B.REQUEST_BATCH_ID
IN
(
SELECT REQUEST_BATCH_ID
FROM bpdata.bulk_prov_master A
where A.START_TIME > to_DATE ('12/16/2014 00', 'MM/DD/YYYY HH24')
)
GROUP BY To_CHAR(B.UPDATE_STAMP , 'DD/MM/YYYY HH24'),REQUEST_BATCH_ID
order by 1 desc,2 desc
and I get the data in following format
Batch_ID Day/Date Count(*)
257658 17/12/2014 11 5
257658 17/12/2014 12 4
257657 17/12/2014 12 4
257656 17/12/2014 12 2
But now I want to count the total number of count() and the largest number of count() and I am unable to devise a query for that. Your help and guidance would be appreciated in that regard
EDITED:
Example of Required Output
Batch_ID Day/Date Count(*)
257658 17/12/2014 11 5
257658 17/12/2014 12 4
257657 17/12/2014 12 4
257656 17/12/2014 12 2
Sum(Count(*)) Largest_request Time
15 5 17/12/2014 11
Sum(Count(*)) and Largest_request and time is what I need now

Answer 1

I haven't tested, but it should be something like:
SELECT SUM(x.total) AS grandtotal
FROM (
SELECT REQUEST_BATCH_ID,To_CHAR(B.UPDATE_STAMP , 'DD/MM/YYYY HH24'),COUNT(*) AS total
FROM bpdata.bulk_prov_detail B
WHERE B.REQUEST_BATCH_ID
IN
(
SELECT REQUEST_BATCH_ID
FROM bpdata.bulk_prov_master A
where A.START_TIME > to_DATE ('12/16/2014 00', 'MM/DD/YYYY HH24')
)
GROUP BY To_CHAR(B.UPDATE_STAMP , 'DD/MM/YYYY HH24'),REQUEST_BATCH_ID
) x
FOR Max Count:
SELECT *
FROM (
SELECT REQUEST_BATCH_ID,To_CHAR(B.UPDATE_STAMP , 'DD/MM/YYYY HH24'),COUNT(*) AS total
FROM bpdata.bulk_prov_detail B
WHERE B.REQUEST_BATCH_ID
IN
(
SELECT REQUEST_BATCH_ID
FROM bpdata.bulk_prov_master A
where A.START_TIME > to_DATE ('12/16/2014 00', 'MM/DD/YYYY HH24')
)
GROUP BY To_CHAR(B.UPDATE_STAMP , 'DD/MM/YYYY HH24'),REQUEST_BATCH_ID
) x ORDER BY x.total DESC LIMIT 0,1
To get exactly what you are asking for, you join the 2 queries. (also untested)
SELECT
(
SELECT SUM(x.total) AS grandtotal
FROM (
SELECT REQUEST_BATCH_ID,To_CHAR(B.UPDATE_STAMP , 'DD/MM/YYYY HH24'),COUNT(*) AS total
FROM bpdata.bulk_prov_detail B
WHERE B.REQUEST_BATCH_ID
IN
(
SELECT REQUEST_BATCH_ID
FROM bpdata.bulk_prov_master A
where A.START_TIME > to_DATE ('12/16/2014 00', 'MM/DD/YYYY HH24')
)
GROUP BY To_CHAR(B.UPDATE_STAMP , 'DD/MM/YYYY HH24'),REQUEST_BATCH_ID
) x
) y,
(
SELECT *
FROM (
SELECT To_CHAR(B.UPDATE_STAMP , 'DD/MM/YYYY HH24') AS daydate,COUNT(*) AS total
FROM bpdata.bulk_prov_detail B
WHERE B.REQUEST_BATCH_ID
IN
(
SELECT REQUEST_BATCH_ID
FROM bpdata.bulk_prov_master A
where A.START_TIME > to_DATE ('12/16/2014 00', 'MM/DD/YYYY HH24')
)
GROUP BY To_CHAR(B.UPDATE_STAMP , 'DD/MM/YYYY HH24'),REQUEST_BATCH_ID
) x ORDER BY x.total DESC LIMIT 0,1
) z

Answer 2

You can use Subquery Factoring WITH Clause to achieve the same
with batch_query as
(
SELECT REQUEST_BATCH_ID,
To_CHAR(B.UPDATE_STAMP, 'DD/MM/YYYY HH24') upd_stmp,
COUNT(*) cnt
FROM bpdata.bulk_prov_detail B
WHERE B.REQUEST_BATCH_ID IN
(SELECT REQUEST_BATCH_ID
FROM bpdata.bulk_prov_master A
where A.START_TIME >
to_DATE('12/16/2014 00', 'MM/DD/YYYY HH24'))
GROUP BY To_CHAR(B.UPDATE_STAMP, 'DD/MM/YYYY HH24'),
REQUEST_BATCH_ID)
,batch_agg_query as
(SELECT max(cnt) max_cnt,sum(cnt) sum_cnt from batch_query )
select sum_cnt,max_cnt,batch_query.upd_stmp
FROM batch_query,batch_agg_query
where batch_query.cnt=batch_agg_query.max_cnt