diff options
| author | Alvaro Herrera <alvherre@alvh.no-ip.org> | 2020-10-14 20:12:26 -0300 |
|---|---|---|
| committer | Alvaro Herrera <alvherre@alvh.no-ip.org> | 2020-10-14 20:12:26 -0300 |
| commit | 4e9821b6fac5042e872d5397f711a67984b165f8 (patch) | |
| tree | 805886d73e19331696339de5aa94e6cdefc3808b /doc/src | |
| parent | b94109ce375b137f235149bfba3559c69f4573e7 (diff) | |
| download | postgresql-4e9821b6fac5042e872d5397f711a67984b165f8.tar.gz postgresql-4e9821b6fac5042e872d5397f711a67984b165f8.zip | |
Restore replication protocol's duplicate command tags
I removed the duplicate command tags for START_REPLICATION inadvertently
in commit 07082b08cc5d, but the replication protocol requires them. The
fact that the replication protocol was broken was not noticed because
all our test cases use an optimized code path that exits early, failing
to verify that the behavior is correct for non-optimized cases. Put
them back.
Also document this protocol quirk.
Add a test case that shows the failure. It might still succeed even
without the patch when run on a fast enough server, but it suffices to
show the bug in enough cases that it would be noticed in buildfarm.
Author: Álvaro Herrera <alvherre@alvh.no-ip.org>
Reported-by: Henry Hinze <henry.hinze@gmail.com>
Reviewed-by: Petr Jelínek <petr.jelinek@2ndquadrant.com>
Discussion: https://postgr.es/m/16643-eaadeb2a1a58d28c@postgresql.org
Diffstat (limited to 'doc/src')
| -rw-r--r-- | doc/src/sgml/protocol.sgml | 8 |
1 files changed, 5 insertions, 3 deletions
diff --git a/doc/src/sgml/protocol.sgml b/doc/src/sgml/protocol.sgml index f5e33181061..5e06c7523d8 100644 --- a/doc/src/sgml/protocol.sgml +++ b/doc/src/sgml/protocol.sgml @@ -2059,8 +2059,9 @@ The commands accepted in replication mode are: the switch position is the end of the WAL that was streamed, but there are corner cases where the server can send some WAL from the old timeline that it has not itself replayed before promoting. Finally, the - server sends CommandComplete message, and is ready to accept a new - command. + server sends two CommandComplete messages (one that ends the CopyData + and the other ends the <literal>START_REPLICATION</literal> itself), and + is ready to accept a new command. </para> <para> @@ -2382,7 +2383,8 @@ The commands accepted in replication mode are: <para> The messages inside the CopyBothResponse messages are of the same format - documented for <literal>START_REPLICATION ... PHYSICAL</literal>. + documented for <literal>START_REPLICATION ... PHYSICAL</literal>, including + two CommandComplete messages. </para> <para> |