"Ed L." <pggeneral@bluepolka.net> writes:
> On Monday February 17 2003 5:09, Tom Lane wrote:
>> Hm, I suddenly have a theory. Do you guys use LISTEN/NOTIFY in the
>> databases that are exhibiting the problem?
> Not using LISTEN/NOTIFY here, at least not of which I'm aware. I'll
> dig a little more and let you know if I find differently.
Oh, duh! You don't need to be using NOTIFY explicitly --- you just have
to sit idle long enough to make the SI buffer overflow, and the system
will try to NOTIFY your backend anyway to make it read the SI message
buffer. Which is the path I was looking at --- that will leave
ImmediateInterruptOK false, preventing response to SIGTERM if the client
still hasn't done anything by the time it comes.
Patch forthcoming.
regards, tom lane