Hello all, We recently created a 11g database on Red Hat Linux 5 and are currently bechmarking our ETLs. I'm noticing the following message in the alert log quite often. "Active Session History (ASH) performed an emergency flush" I looked it up on metalink and the recommendation is to increase "_ash_size". Listed below is the size on our server. Its 134 MB. select total_size,awr_flush_emergency_count from v$ash_info; TOTAL_SIZE AWR_FLUSH_EMERGENCY_COUNT ------------------------- 134217728 20 Have you encountered this issue? Has anyone bumped up the size of the hidden parameter "_ash_size" to resolve this issue. I normally try to avoid setting hidden parameters as much as possible. Thanks for your time.
I would suggest getting touch with Oracle support, confirming that yours is the same case as reported in the MOS document and then ask support only to provide a recommendation of value to be set.