Tuning Recovery - Expert Oracle RAC Performance Diagnostics and Tuning - page 322

to_char((sum(decode(to_char(first_time,'HH24'),'00',1,0))+
sum(decode(to_char(first_time,'HH24'),'01',1,0))+
sum(decode(to_char(first_time,'HH24'),'02',1,0))+
sum(decode(to_char(first_time,'HH24'),'03',1,0))+
sum(decode(to_char(first_time,'HH24'),'04',1,0))+
sum(decode(to_char(first_time,'HH24'),'05',1,0))+
sum(decode(to_char(first_time,'HH24'),'06',1,0))+
sum(decode(to_char(first_time,'HH24'),'07',1,0))+
sum(decode(to_char(first_time,'HH24'),'08',1,0))+
sum(decode(to_char(first_time,'HH24'),'09',1,0))+
sum(decode(to_char(first_time,'HH24'),'10',1,0))+
sum(decode(to_char(first_time,'HH24'),'11',1,0))+
sum(decode(to_char(first_time,'HH24'),'12',1,0))+
sum(decode(to_char(first_time,'HH24'),'13',1,0))+
sum(decode(to_char(first_time,'HH24'),'14',1,0))+
sum(decode(to_char(first_time,'HH24'),'15',1,0))+
sum(decode(to_char(first_time,'HH24'),'16',1,0))+
sum(decode(to_char(first_time,'HH24'),'17',1,0))+
sum(decode(to_char(first_time,'HH24'),'18',1,0))+
sum(decode(to_char(first_time,'HH24'),'19',1,0))+
sum(decode(to_char(first_time,'HH24'),'20',1,0))+
sum(decode(to_char(first_time,'HH24'),'21',1,0))+
sum(decode(to_char(first_time,'HH24'),'22',1,0))+
sum(decode(to_char(first_time,'HH24'),'23',1,0))),'999') total
from v$log
where trunc(first_time) > (trunc(sysdate) - 7)
group by substr(to_char(first_time,'MM/DD-Day'),1,9)
order by 1 ;
It is noticed in the output and the graphical view of the data (Figure 10-1 ) that there are high redo log switch rates
throughout the day. However, it's extremely high between 2 AM and 10 AM every day. Further analysis of the data
revealed that during this period, the application has high batch load activity. Comparing this to the ideal switch rates
of four or five per hour, the current redo log switch intervals are significantly high.