sql register entries (within group)
sql register entries (within group)
i have little problem i'm certain best proceed repair it, i wholly have unparalleled opening database (oracle) itself.
in the list "event" have 160k entries, any eventuality groupid normal opening accurately 5 rows same groupid. due bug now integrate register entries (duplicate, 10 rows instead 5, only opposite eventid. competence change, it's only <> 5). need filter entries groups.
due unparalleled opening database surrogate table, nor supplement an index groupid buttress faster.
we groupids query, nonetheless need second query indispensable data
select a."groupid"
from "event" a
group a."groupid"
having count(a."groupid") <> 5
one fortitude subselect:
select *
from "event" a
where a."groupid" (
name b."groupid"
"event" b
organisation b."groupid"
carrying count(b."groupid") <> 5
)
without an index groupid 160k entries, takes many too long.
tried pondering join hoop this, nonetheless can't good fortitude far.
anybody good fortitude maybe?
small edit:
we don't have 100% duplicates here, any opening still unparalleled id groupid unparalleled presumably (that's since need "group by") - maybe i only skip an easy fortitude :)
small instance information (i don't wish mislay it, only it)
eventid | groupid | typeid
123456 123 12
123457 123 145
123458 123 2612
123459 123 41
123460 123 238
234567 123 12
234568 123 145
234569 123 2612
234570 123 41
234571 123 238
it some-more columns, timestamp etc, nonetheless already, all identical, besides eventid.
we run some-more mostly testing, bug check happens again.
Comments
Post a Comment