Re: wal_dump output on CREATE DATABASE
От | Andres Freund |
---|---|
Тема | Re: wal_dump output on CREATE DATABASE |
Дата | |
Msg-id | 20181113202459.diyrpnw3tqa474wb@alap3.anarazel.de обсуждение исходный текст |
Ответ на | Re: wal_dump output on CREATE DATABASE (Alvaro Herrera <alvherre@2ndquadrant.com>) |
Ответы |
Re: wal_dump output on CREATE DATABASE
|
Список | pgsql-hackers |
Hi, On 2018-11-13 17:18:32 -0300, Alvaro Herrera wrote: > On 2018-Nov-13, Peter Eisentraut wrote: > > > On 13/11/2018 18:53, Jean-Christophe Arnu wrote: > > > May I request any update on my (not so important) proposal ? > > > > Other opinions out there perhaps? I think your proposals are good, but > > I'm not sure to what extent people are automatically parsing the > > pg_waldump output and are relying on a particular fixed format. I personally don't care either way about this change. But: > While I've used pg_waldump a number of times to investigate various > problems, I have never written a script that would be broken by the > proposed changes. It seems the requirements vary every time. I'm not too concerned either. There's plenty change of the WAL contents across versions anyway. We shouldn't break things gratuitously, but we shouldn't hesitate to make the format better either. If somebody really wanted something that parses reasonably across versions I'd like to a) hear that usecase b) come up with an output format that's oriented towards that. Greetings, Andres Freund
В списке pgsql-hackers по дате отправления: