Database Business Problems at Oracle?
Sorry for the delayed reply - been busy this past week. This is under RH9.2. I was not aware of an out of memory issue with hashed aggregates, so I suppose that could be related... what would be the easiest way to check? I have noticed though that any type of backend crash under Linux results in all the other backends being told to restart due to possible corrupted shared memory... does that not occur under other OS'?