Re: [HACKERS] pg_dump vs. idle_in_transaction_session_timeout

2016-06-15 Thread Tom Lane
Bernd Helmle writes: > Currently pg_dump doesn't turn off idle_in_transaction_session_timeout. Definitely an oversight, since it turns off other timeout settings. > Okay, the window of failure here is very narrow (on my machine it breaks > with an insane setting of 1ms only), Probably if you we

[HACKERS] pg_dump vs. idle_in_transaction_session_timeout

2016-06-15 Thread Bernd Helmle
Currently pg_dump doesn't turn off idle_in_transaction_session_timeout. Okay, the window of failure here is very narrow (on my machine it breaks with an insane setting of 1ms only), but for the sake of reliable backups and protection against over motivated DBA it looks better to me to turn that of