Bad Plan picked up
Hi All,
I would request your suggestion on this as one of my query picked up bad plan ended in long run. This job usually runs for 10mins everyday but today it ran more than 3hrs and we had to terminate, pin the right plan which ran in time afterwards. Also no changes were made to this program or table data and gather stats are upto date.
We have 8 node rac exadata , the bad plan generated on 2nd instance and good plan was in 6th instance.
Any idea why the 2nd instance optimizer wants to choose different plan as below?