Stale statistics and rows appended to the end of the table
Found this statement in one of the documents attached to "pscbo_stats - Improving Statistics in Oracle RDBMS for PeopleSoft Enterprise [ID 1322888.1]"
"Since ERP applications tend to localize modifications at the the “end” of a table, it is possible that over time the overall statistics don’t reflect the actually data well, even though they are not flagged as “stale”."
PeopleSoft or not, does this make sense to anyone? I thought it was 10% no matter what modification was and where it occured. Is that not correct?
"Since ERP applications tend to localize modifications at the the “end” of a table, it is possible that over time the overall statistics don’t reflect the actually data well, even though they are not flagged as “stale”."
PeopleSoft or not, does this make sense to anyone? I thought it was 10% no matter what modification was and where it occured. Is that not correct?
0