Re: Make check problem with 7.2
От | Tom Lane |
---|---|
Тема | Re: Make check problem with 7.2 |
Дата | |
Msg-id | 24237.1015208120@sss.pgh.pa.us обсуждение исходный текст |
Ответ на | Make check problem with 7.2 (Steven Núñez <nunez@itl-global.com>) |
Ответы |
Re: Make check problem with 7.2
|
Список | pgsql-bugs |
Steven Núñez <nunez@itl-global.com> writes: > This time I've included the log file... > DEBUG: connection startup failed (fork failure): Resource temporarily > unavailable > DEBUG: connection startup failed (fork failure): Resource temporarily > unavailable > DEBUG: connection startup failed (fork failure): Resource temporarily > unavailable > DEBUG: connection startup failed (fork failure): Resource temporarily > unavailable > DEBUG: connection startup failed (fork failure): Resource temporarily > unavailable > DEBUG: connection startup failed (fork failure): Resource temporarily > unavailable At a guess, you have the max-processes-per-user limit set too low. The postmaster attempts to report the fork failure to the client, but it looks like (at least on your platform) libpq gives up before receiving the error message. I wonder if it'd make sense for libpq to ignore send failure on the startup packet, so it could move ahead to receive the error message. I have a feeling that would make the behavior worse in other scenarios, though, so it may not be a win. regards, tom lane
В списке pgsql-bugs по дате отправления: