Re: background worker shudown (was Re: [HACKERS] Why does logical replication launcher exit with exit code 1?)

2018-10-10 Thread Robert Haas
On Wed, Oct 10, 2018 at 12:00 AM Thomas Munro wrote: > So I suppose we should just remove it, with something like 0002. I'm > a bit uneasy about existing code out there that might be not calling > CFI. OTOH I suspect that a lot of code copied worker_spi.c and > installed its own handler. I agre

Re: background worker shudown (was Re: [HACKERS] Why does logical replication launcher exit with exit code 1?)

2018-10-09 Thread Thomas Munro
On Wed, Oct 10, 2018 at 7:29 AM Robert Haas wrote: > On Thu, Oct 4, 2018 at 7:37 PM Thomas Munro > wrote: > > 0. The default SIGTERM handler for bgworkers is bgworker_die(), which > > generates a FATAL ereport "terminating background worker \"%s\" due to > > administrator command", directly in t

Fwd: background worker shudown (was Re: [HACKERS] Why does logical replication launcher exit with exit code 1?)

2018-10-09 Thread Robert Haas
On Thu, Oct 4, 2018 at 7:37 PM Thomas Munro wrote: > I still think the current situation is non-ideal. I don't have a > strong view on whether some or all system-wide processes should say > hello and goodbye explicitly in the log, but I do think we need a way > to make that not look like an error