Commit Graph

22 Commits

Author SHA1 Message Date
Alexander Kukushkin
6cff5a3e08 Smart replication feedback
This commit makes psycopg2 responsible for sending the status update
(feedback) messages to the server regardless of whether a synchronous or
asynchronous connection is used.

Feedback is sent every *status_update* (default value is 10) seconds,
which could be configured by passing a corresponding parameter to the
`start_replication()` or `start_replication_expert()` methods.
The actual feedback message is sent by the
`pq_read_replication_message()` when the *status_update* timeout is
reached.

The default behavior of the `send_feedback()` method is changed.
It doesn't send a feedback message on every call anymore but just
updates internal structures. There is still a way to *force* sending
a message if *force* or *reply* parameters are set.

The new approach has certain advantages:
1. The client can simply call the `send_feedback()` for every
   processed message and the library will take care of not overwhelming
   the server. Actually, in the synchronous mode it is even mandatory
   to confirm every processed message.
2. The library tracks internally the pointer of the last received
   message which is not keepalive. If the client confirmed the last
   message and after that server sends only keepalives with increasing
   *wal_end*, the library can safely move forward *flush* position to
   the *wal_end* and later automatically report it to the server.

Reporting of the *wal_end* received from keepalive messages is very
important. Not doing so casing:
1. Excessive disk usage, because the replication slot prevents from
   WAL being cleaned up.
2. The smart and fast shutdown of the server could last indefinitely
   because walsender waits until the client report *flush* position
   equal to the *wal_end*.

This implementation is only extending the existing API and therefore
should not break any of the existing code.
2019-05-06 10:27:44 +02:00
Martins Grunskis
f946042a79 Store WAL end pointer in the replication cursor 2019-03-30 21:23:11 +00:00
Daniele Varrazzo
b0119fef81 Other import moved to top level in tests 2019-03-16 19:13:58 +00:00
Jon Dufresne
d90ad8627d Move imports to the top of the module across tests
Allows removing many duplicate imports and better follows PEP8
guidelines:

https://www.python.org/dev/peps/pep-0008/#imports

> Imports are always put at the top of the file, just after any module
> comments and docstrings, and before module globals and constants.
2019-03-16 18:03:41 +00:00
Daniele Varrazzo
599432552a Copyright year updated
ag -l Copyright | xargs sed -i \
        "s/\(.*copyright (C) [0-9]\+\)\(-[0-9]\+\)\?\(.*\)/\1-$(date +%Y)\3/I"
2019-02-17 01:36:36 +00:00
Jon Dufresne
1a8e992fcc Use relative imports throughout tests
The tests relied on Python2 relative import semantics. Python3 changed
import semantics to always search sys.path by default. To import using a
relative path it must have a leading dot.

Forward compatible with newer Pythons.

Works towards the goal of moving tests outside of the installed package.

For more information, see PEP-328:

https://www.python.org/dev/peps/pep-0328/
2017-12-10 10:51:07 -08:00
Jon Dufresne
fcc083dc12 Always import the system unittest
There is no need to import testutils.unittest instead of simply
unittest. They are simple aliases. Use system unittest to be more
regular, consistent as well as idiomatic with the wider Python
community.
2017-12-01 18:35:30 -08:00
Daniele Varrazzo
1a97445471 Accept Composable in start_replication_expert()
Close #554
2017-06-17 03:30:44 +01:00
Daniele Varrazzo
248e653c9e Fixed args parsing in ReplicationCursor.consume_stream()
Close #547.
2017-04-19 01:01:59 +01:00
Daniele Varrazzo
ce9be69615 Added async_ as an alias for async
Added in argument for psycopg2.connect() and connection.__init__, and
for the connection.async attribute.
2017-02-03 04:43:49 +00:00
Daniele Varrazzo
c22093ddd4 Skip replication tests in green mode 2016-12-25 19:00:30 +01:00
Daniele Varrazzo
b73115ac41 Added test to verify bug #482 2016-12-25 18:30:27 +01:00
Daniele Varrazzo
a478ba9a47 Fixed tests failing on Python 2.6 2016-12-24 01:03:57 +01:00
Daniele Varrazzo
91d2158de7 Python source cleanup using flake8 2016-10-11 00:11:55 +01:00
Oleksandr Shulgin
5d33b39829 Fix error test for invalid START_REPLICATION command. 2016-01-21 15:56:27 +01:00
Oleksandr Shulgin
09a4bb70a1 Allow retrying start_replication after syntax or data error. 2016-01-05 12:31:57 +01:00
Oleksandr Shulgin
8b79bf43ac Drop ReplicationCursor.flush_feedback(), rectify pq_*_replication_*() interface. 2015-10-23 18:30:18 +02:00
Oleksandr Shulgin
0bb81fc848 Properly subclass ReplicationCursor on C level. 2015-10-19 20:00:39 +02:00
Oleksandr Shulgin
7aea2cef6e Improve async replication test. 2015-10-19 17:02:18 +02:00
Oleksandr Shulgin
4ab7cf0157 Replace stop_replication with requirement for an exception. 2015-10-19 15:42:42 +02:00
Oleksandr Shulgin
0435320f34 Fix PSYCOPG2_TEST_REPL_DSN handling. 2015-10-16 16:36:03 +02:00
Oleksandr Shulgin
cf4f2411bf Fix async replication and test. 2015-10-15 18:01:43 +02:00