Re: Results change when embedding sub-queries

[Rick James]

    select  count(ext.stbId), count(inv.stbId)
        from  
          ( SELECT  *
                from  aaa.extendeddata
                where  aaa.extendeddata.stbId in ('R1881445299')
          ) as ext,
          ( SELECT  *
                from  aaa.invidifilelist
                where  aaa.invidifilelist.stbId in ('R1881445299')
          ) as inv;

Use the JOIN ... ON syntax.

Don't specify anything in COUNT(...) unless you are trying to avoid NULLs. Simply say COUNT(*).

In newer versions, IN(single-item) is just as fast as = single item.

Perhaps this is what you want?

    select 
        ( SELECT COUNT(*)
                from  aaa.extendeddata
                where  aaa.extendeddata.stbId in ('R1881445299')
        ) as ext,
        ( SELECT COUNT(*)
                from  aaa.invidifilelist
                where  aaa.invidifilelist.stbId in ('R1881445299')
        ) as inv;

It gives one row with 2 columns, the counts from each table. And it is reasonably efficient.