Re: Transaction ISOLATION LEVEL - have I missed something?
От | David G. Johnston |
---|---|
Тема | Re: Transaction ISOLATION LEVEL - have I missed something? |
Дата | |
Msg-id | CAKFQuwbr1ycMQjytczn4KBhR85XdjAB-TUBDTANiPKerURL_Aw@mail.gmail.com обсуждение исходный текст |
Ответ на | Re: Transaction ISOLATION LEVEL - have I missed something? (Pól Ua Laoínecháin <linehanp@tcd.ie>) |
Ответы |
Re: Transaction ISOLATION LEVEL - have I missed something?
|
Список | pgsql-novice |
On Thursday, May 13, 2021, Pól Ua Laoínecháin <linehanp@tcd.ie> wrote:
Hi David, and thanks for getting back to me.
>> Am I wrong about this?
> Transaction isolation levels are immaterial here. The functions define what time they return:
Ah, yes, it now makes perfect sense...
All statements in a transaction which is READ COMMITTED sees the data
as of CLOCK_TIMESTAMP() and all statements in a transaction that is
SERIALIZABLE see the data as of TRANSACTION_TIMESTAMP().
At least, that's how I understand it now - my error was not
double-checking my function definitions.
No. The isolation level is immaterial to which times are returned. I welcome being proven wrong but nothing I’ve read in that link suggests otherwise so I’m disinclined to prove it correct.
David J.
В списке pgsql-novice по дате отправления: