Richard Huxton <dev@xxxxxxxxxxxx> writes: > Tom Lane wrote: >> FWIW, a look in the source code shows that the 'corrupted item pointer' >> message comes only from PageIndexTupleDelete, so that indicates a >> damaged index which should be fixable by reindexing. > Tom - could it be damage to a shared system-catalogue, and template1 > just the first DB to be vacuumed? Possible, but in any case the other error indicates an independent problem. > It just strikes me as odd that an > index on template1 would be corrupted - assuming it's your typical empty > template1 and is just being connected to during DB creation etc. Yeah, I was wondering about that too. PageIndexTupleDelete wouldn't even get called unless there was something to delete, which indicates a table that's been changed. Is the OP in the habit of doing real work in template1? regards, tom lane
