Since deadlock it's produced in a trigger which you said it's kind of big, maybe this scenario happens:
2 transactions running concurently:isc_tpb_read_committed + isc_tpb_no_rec_version + wait Transaction1 locks(in trigger) first some TABLE1 and want to lock TABLE2 Transaction2 locks(in the same trigger but maybe with some IF statement) first TABLE2 and the going to lock TABLE1. You have nice DEADLOCK, each transaction waitting for each other. ---In [email protected], <nathanelrick@...> wrote: no one have an idea about what is a deadlock Error: 16 ? the isolation of the transaction was: isc_tpb_read_committed + isc_tpb_no_rec_version + wait => normally no deadlock must appear ?? --- In [email protected] mailto:[email protected], "nathanelrick" <nathanelrick@...> wrote: > > hello, > > what is the most fastest isolation level ? > i know the behavior of each, but i need to know the difference in speed / > resource usage between each of them ... > > isc_tpb_concurrency > isc_tpb_consistency > isc_tpb_read_committed + isc_tpb_rec_version > isc_tpb_read_committed + isc_tpb_no_rec_version > > thanks by advance >
