Re: [GENERAL] session_replication_role `replica` behavior

2013-04-25 Thread Achilleas Mantzios
Point taken, thanx, however from the docs, it is far from explicit that setting session_replication_role to 'replica' can disable FK constraints (RI) and finally result in an incosistent database. It might be that RI in postgres is implemented via triggers, but to the user, that is just an implem

Re: [GENERAL] session_replication_role `replica` behavior

2013-04-25 Thread Greg Sabino Mullane
-BEGIN PGP SIGNED MESSAGE- Hash: RIPEMD160 manos tsahakis wrote: > In our application we are enabling session_replication_role TO 'replica' in > certain situations so that triggers will not fire in a table during DML > operations. However, we observed that when setting session_replicatio

[GENERAL] session_replication_role 'replica' behavior

2013-04-24 Thread Manos Tsahakis
Hello all, In our application we are enabling session_replication_role TO 'replica' in certain situations so that triggers will not fire in a table during DML operations. However, we observed that when setting session_replication_role TO 'replica' referential integrity constraints will not fire on