Hello , Some new input: On slave, all domains ( with checks) have been replaced by a simple type. No crash on slave since this bypass. Is there something to fix in the ActiveSnapshot code ? BR
> Le 18 juil. 2018 à 17:03, Tom Lane <t...@sss.pgh.pa.us> a écrit : > > Mai Peng <maily.p...@webedia-group.com> writes: >> Here the backtrace > > Hmm .. so this can be summarized as "logical replication workers should > provide an ActiveSnapshot in case the user functions they call want one". > Makes me wonder how much other transactional infrastructure is needed > but not present. > > regards, tom lane