Re: Change pg_cancel_*() to ignore current backend
От | Jon Nelson |
---|---|
Тема | Re: Change pg_cancel_*() to ignore current backend |
Дата | |
Msg-id | CAKuK5J12ynF2k4WsL7VBOHHwN-idUEohqYYZx=pwUZEkAQ0T9A@mail.gmail.com обсуждение исходный текст |
Ответ на | Re: Change pg_cancel_*() to ignore current backend (David Steele <david@pgmasters.net>) |
Список | pgsql-hackers |
<p dir="ltr"><br /> On May 20, 2015 6:43 AM, "David Steele" <<a href="mailto:david@pgmasters.net">david@pgmasters.net</a>>wrote:<br /> ><br /> > On 5/20/15 1:40 AM, Jim Nasby wrote:<br/> > > On 5/19/15 9:19 PM, Fabrízio de Royes Mello wrote:<br /> > >> We could add a second parameterto the current functions:<br /> > >> allow_own_pid DEFAULT false. To me that seems better than an<br/> > >> entirely separate set of functions.<br /> > >><br /> > >><br /> > >>+1 to add a second parameter to current functions.<br /> > ><br /> > > Instead of allow_own_pid, I wentwith skip_own_pid. I have the function<br /> > > still returning true even when it skips it's own PID... that seemsa bit<br /> > > weird, but I think it's better than returning false. Unless someone<br /> > > thinks itshould return NULL, but I don't see that as any better either.<br /> ><br /> > +1. I agree that cancelling/killingyour own process should not be the<br /> > default behavior.<br /><p dir="ltr">-1. It breaks backwardscompatibility. I use this function a fair bit to terminate the current backend all the time. Changing the signatureis great, by make the default behavior retain the current behavior.<p dir="ltr">><br /> > --<br /> > -David Steele<br /> > <a href="mailto:david@pgmasters.net">david@pgmasters.net</a><br /> ><br />
В списке pgsql-hackers по дате отправления: