Re: logical decoding bug when mapped relation with toast contents is rewritten repeatedly

2018-10-10 Thread Andres Freund
Hi, On 2018-09-13 19:10:46 -0700, Andres Freund wrote: > (Tomas, CCing you because you IIRC mentioned encountered an issue like > this) > > I just spent quite a while debugging an issue where running logical > decoding yielded a: > ERROR: could not map filenode "base/$X/$Y" to relation OID > err

Re: logical decoding bug when mapped relation with toast contents is rewritten repeatedly

2018-09-20 Thread Andres Freund
Hi, On 2018-09-14 16:13:46 +0200, Tomas Vondra wrote: > > > > I suspect the proper fix would be to have a new HEAP_INSERT_NO_LOGICAL > > option, and specify that in raw_heap_insert() iff > > RelationIsLogicallyLogged(state->rs_old_rel) or something like that. > > > > Attached is a *prototype* pa

Re: logical decoding bug when mapped relation with toast contents is rewritten repeatedly

2018-09-14 Thread Tomas Vondra
Hi, On 09/14/2018 04:10 AM, Andres Freund wrote: > Hi, > > (Tomas, CCing you because you IIRC mentioned encountered an issue like > this) > I might have mentioned an issue with this symptom recently, but that turned out to be already fixed by da10d6a8a9 (before the minor version with that fix g

logical decoding bug when mapped relation with toast contents is rewritten repeatedly

2018-09-13 Thread Andres Freund
Hi, (Tomas, CCing you because you IIRC mentioned encountered an issue like this) I just spent quite a while debugging an issue where running logical decoding yielded a: ERROR: could not map filenode "base/$X/$Y" to relation OID error. After discarding like 30 different theories, I have found th