I am running the below query on cluster env:

select
sid,
username,
round(100 * total_user_io/total_io,2) tot_io_pct
from
(select
b.sid sid,
nvl(b.username,p.name) username,
sum(value) total_user_io
from
sys.gv_$statname c,
sys.gv_$sesstat a,
sys.gv_$session b,
sys.gv_$bgprocess p
where
a.statistic#=c.statistic# and
p.paddr (+) = b.paddr and
b.sid=a.sid and
c.name in ('physical reads',
'physical writes',
'physical writes direct',
'physical reads direct',
'physical writes direct (lob)',
'physical reads direct (lob)')
group by
b.sid, nvl(b.username,p.name)),
(select
sum(value) total_io
from
sys.gv_$statname c,
sys.gv_$sesstat a
where
a.statistic#=c.statistic# and
c.name in ('physical reads',
'physical writes',
'physical writes direct',
'physical reads direct',
'physical writes direct (lob)',
'physical reads direct (lob)'))
order by
3 desc;


I got a huge amount of total_io_pct on DBW0 (over 300), what could cause this huge number for DBW0?

thanks,