Re: [HACKERS] Walsender timeouts and large transactions
Kyotaro HORIGUCHI <horiguchi.kyotaro@lab.ntt.co.jp>
From: Kyotaro HORIGUCHI <horiguchi.kyotaro@lab.ntt.co.jp>
To: petr.jelinek@2ndquadrant.com
Cc: robertmhaas@gmail.com, funny.falcon@postgrespro.ru,
pgsql-hackers@postgresql.org, pjmodos@pjmodos.net
Date: 2017-11-17T04:35:55Z
Lists: pgsql-hackers
Hello, At Fri, 17 Nov 2017 13:24:08 +0900 (Tokyo Standard Time), Kyotaro HORIGUCHI <horiguchi.kyotaro@lab.ntt.co.jp> wrote in <20171117.132408.85564852.horiguchi.kyotaro@lab.ntt.co.jp> > > Well, even the CHECK_FOR_INTERRUPTS() can be called unconditionally yes. > > It just seems like it's needless call as we'll call both in for loop > > anyway if we take the "slow" path. I admit it's not exactly big win > > though. If you think it would improve readability I can move it. > > I think this is the last message in this thread so I changed the > status of the CF entry to "Waiting for Author". Hmm. Somehow the last patch and Robert's comment * which is the base of the patch * has been reached me in reverse order. I found that the patch is the latest one and will look this soon. Sorry for the ignorance. regards, -- Kyotaro Horiguchi NTT Open Source Software Center
Commits
-
Fix walsender timeouts when decoding a large transaction
- f5c7e0cddf6a 9.4.16 landed
- 87056267eb85 9.5.11 landed
- c28e0b1e0ab9 9.6.7 landed
- 14c15b1f45c8 10.2 landed
- 0fedb4ea6946 11.0 landed