Re: [HACKERS] make async slave to wait for lsn to be replayed

Intro
==========
The main purpose of the feature is to achieve
read-your-writes-consistency, while using async replica for reads and
primary for writes. In that case lsn of last modification is stored
inside application. We cannot store this lsn inside database, since
reads are distributed across all replicas and primary.

Two implementations of one feature
==========
We left two different solutions. Help me please to choose the best.

1) Classic (wait_classic_v7.patch)
https://www.postgresql.org/message-id/3cc883048264c2e9af022033925ff8db%40postgrespro.ru
Synopsis
==========
advantages: multiple wait events, separate WAIT FOR statement
disadvantages: new words in grammar

WAIT FOR [ANY | ALL] event [, ...]
BEGIN [ WORK | TRANSACTION ] [ transaction_mode [, ...] ]
[ WAIT FOR [ANY | ALL] event [, ...]]
event:
LSN value
TIMEOUT number_of_milliseconds
timestamp

2) After style: Kyotaro and Freund (wait_after_within_v6.patch)
https://www.postgresql.org/message-id/d3ff2e363af60b345f82396992595a03%40postgrespro.ru
Synopsis
==========
advantages: no new words in grammar
disadvantages: a little harder to understand, fewer events to wait

AFTER lsn_event [ WITHIN delay_milliseconds ] [, ...]
BEGIN [ WORK | TRANSACTION ] [ transaction_mode [, ...] ]
[ AFTER lsn_event [ WITHIN delay_milliseconds ]]
START [ WORK | TRANSACTION ] [ transaction_mode [, ...] ]
[ AFTER lsn_event [ WITHIN delay_milliseconds ]]

Examples
==========

primary standby
------- --------
postgresql.conf
recovery_min_apply_delay = 10s

CREATE TABLE tbl AS SELECT generate_series(1,10) AS a;
INSERT INTO tbl VALUES (generate_series(11, 20));
SELECT pg_current_wal_lsn();

BEGIN WAIT FOR LSN '0/3002AE8';
SELECT * FROM tbl; // read fresh insertions
COMMIT;

Rebased and ready for review.

--
Ivan Kartyshov
Postgres Professional: www.postgrespro.com