-- Current AWR report (30 Minutes). In addition, if we can get AWR report  when performance was OK would be of help
-- For the same period when AWR report is collected, collect the data for the following (OnDB system)
    -- vmstat 5    
    -- iostat -xm 5
    -- sar -n DEV 5
-- If using OS watcher : Data for 24 hour period when performance was normal. If not running OS watcher - We need the copy of the files available in /var/log/sa directory (all files)
-- Get the output of the following commands (One time execution )
    -- ps -aef
    -- free -m
    -- cat /proc/cpuinfo
    -- jps -l   (OMS only)
    -- jinfo pid  (from output from jps -l )
    -- cat /etc/oracle-release
    --/sbin/sysctl -a  

In addition, it would help, if we can gather information on when started experiencing the slow performance and changes if any made to the systems. The specific wait event noticed on the DB side indicate contention and we need to look at some of the stats from above list to troubleshoot.