gitops/patroni
1
0
Fork 0
mirror of https://github.com/outbackdingo/patroni.git synced 2026-08-31 16:49:46 +00:00
Code Issues Packages Projects Releases Wiki Activity
249 Commits 15 Branches 66 Tags
fb3bf799e90d6c8b5818ea6ce521460636520e21
Commit Graph
146 Commits
Author SHA1 Message Date
Oleksii Kliukin fb3bf799e9 Autodetect the host role during start from the existence of recovery.conf. Fix some typos. 2015-06-03 16:18:47 +02:00
Oleksii Kliukin e06cd49deb autodetect the host role during start from the existence of recovery.conf. 2015-06-03 16:13:03 +02:00
Oleksii Kliukin 1d3a946ffb Fix wrong number of arguments for the EBS instance. 2015-06-03 15:52:35 +02:00
Oleksii Kliukin eb3a120db1 add code to set tags on AWS objects (instances and EBS storage), as well as an ability for the postgres state_handler to call a callback when the state is changed, and use the new AWS code in a callback. 2015-06-03 12:56:00 +02:00
Alexander Kukushkin ccdff5ea99 Merge branch 'master' of github.com:CyberDem0n/governor into features/refactoring 
Conflicts:
	helpers/postgresql.py
	tests/test_postgresql.py
 2015-06-01 14:14:53 +02:00
Alexander Kukushkin adcc7ac256 Try to avoid "double" promotion. 
Also check presence of trigger_file on master after promotion when
pg_is_in_recovery() = false and if it is there - remove it.
Plus check presence of trigger_file on a new slave (after running
pg_basebackup) and if it is there - also remove it.
 2015-06-01 13:48:44 +02:00
Alexander Kukushkin 939254021e Code cleanup: do not modify os.environ but pass copy of it to subprocess.call 2015-06-01 11:03:06 +02:00
Alexander Kukushkin f8c3582715 Refactor api: 
Call is_running only when it's not possible to execute query against
governed postgres.
 2015-06-01 10:48:16 +02:00
Alexander Kukushkin b85262637d Bugfix: do not create replication slot for master 2015-06-01 09:20:24 +02:00
Alexander Kukushkin 5fcb19b817 Merge branch 'sleep-sigchld' of github.com:CyberDem0n/governor into features/refactoring 
Conflicts:
	tests/test_governor.py
 2015-05-28 15:55:03 +02:00
Alexander Kukushkin aa486194ae Wrap time.sleep into wrapper function 
In case if it was interrupted by SIGCHLD but scheduled awake time is not
reached it will continue sleeping. For all other signals behaviuor is
not changed.
 2015-05-28 15:47:51 +02:00
Alexander Kukushkin 9971d214fe Merge branch 'restapi' of github.com:CyberDem0n/governor into features/refactoring 
Conflicts:
	governor.py
 2015-05-27 14:54:08 +02:00
Alexander Kukushkin 5a817fad0d Store in etcd endpoint of restapi 
In order to have backward compatibility connection url of restapi is
passed as application_name parameter in connection url of database.
Value in etcd will look like:
postgres://user:passwd@host:5432/postgres?application_name=http://host:8080/governor
Sure, this is a hack, but I guess it worth it.
 2015-05-27 14:46:35 +02:00
Alexander Kukushkin b163a8896b Merge branch 'restapi' of github.com:CyberDem0n/governor into features/refactoring 
Conflicts:
	helpers/postgresql.py
 2015-05-27 11:20:51 +02:00
Alexander Kukushkin 56a9f52a15 Replace os.system with subprocess.call 2015-05-27 11:16:16 +02:00
Alexander Kukushkin 56daec0d6c Api: Bugfix for current_xlog_location in slaves. 2015-05-26 15:53:57 +02:00
Feike Steenbergen ea416d98d7 Api: Bugfix for current_xlog_location in slaves. 2015-05-26 15:40:32 +02:00
Alexander Kukushkin 8cc0aa90b1 Fix typos when passing connection parameters to psycopg2. 2015-05-26 10:39:16 +02:00
Oleksii Kliukin 7217d0b0c1 Fix typos when passing connection parameters to psycopg2. 2015-05-26 10:35:42 +02:00
Alexander Kukushkin bbf1110518 Merge branch 'restapi' of github.com:CyberDem0n/governor into features/refactoring 
Conflicts:
	governor.py
	helpers/postgresql.py
	postgres0.yml
	postgres1.yml
	tests/test_postgresql.py
 2015-05-26 09:56:35 +02:00
Alexander Kukushkin 518ef68a53 Use connect_timeout=3s and statement_timeout=2s everywhere (even for local connections) 2015-05-26 09:19:20 +02:00
Alexander Kukushkin faf295dcd0 Add initialize key into Cluster 2015-05-24 19:26:35 +02:00
Alexander Kukushkin 1096e04343 Bigfix, cursor.closed contains boolean value 2015-05-24 16:00:31 +02:00
Alexander Kukushkin db34e1f458 Write get_postgresql_status exception into log file 2015-05-24 15:57:10 +02:00
Alexander Kukushkin 8dd99170bf Refactor exceptions handling in etcd.py and ha.py 
update_leader can throw EtcdError exception
unused HealthiestMemberError exception is removed
 2015-05-24 15:40:36 +02:00
Alexander Kukushkin 188741de5a Merge branch 'master' of github.com:CyberDem0n/governor into restapi 2015-05-24 09:02:42 +02:00
Alexander Kukushkin ce267d9f1e query method will retry only in case of communication error 2015-05-24 09:01:23 +02:00
Alexander Kukushkin d202f72a42 Added simple web server which shows current state of postgres 
Server always returns json which contains some status of postgres:
* instance type: master/slave
* xlog location for master
* xlog received_location, replayed_location
This server could be used by haproxy
GET / returns 200 if there is postgres master behind governor.
GET /slave returns 200 is there is postgres slave behind governor
In all other cases it returns 503
Currently server always listening on 0.0.0.0:8080, but it should be
configurable. In the next versions this rest api could also report some
status of etcd and could be used by is_healthiest_node method
 2015-05-23 20:29:51 +02:00
Alexander Kukushkin f53c369c69 Connect timeout 5 seconds for psycopg2 
Refactored query and is_healthiest_node methods
query method wont retry more then 3 times if OperationalError raised
is_healthiest_node takes into account that one of the nodes could be
still a master
 2015-05-23 20:21:13 +02:00
Alexander Kukushkin 7f9c10e33f Merge branch 'master' of github.com:CyberDem0n/governor into features/refactoring 2015-05-22 12:27:02 +02:00
Alexander Kukushkin b00846affc Merge branch 'master' of github.com:CyberDem0n/governor into features/refactoring 
Conflicts:
	helpers/postgresql.py
 2015-05-22 12:26:55 +02:00
Alexander Kukushkin 34f9b666cf Demote master when etcd is not accessible 2015-05-22 12:26:19 +02:00
Alexander Kukushkin 09163ac454 If there is no connect_address defined in yml file use address from listen as a fallback 2015-05-22 12:03:43 +02:00
Alexander Kukushkin 62d17d68de Use "member_ttl" with default value=3600 for members 2015-05-22 12:00:07 +02:00
Alexander Kukushkin bb884bac00 Move is_unlocked method from Ha into Cluster class 2015-05-22 11:45:09 +02:00
Alexander Kukushkin 4a4d9b85e5 Make sure that there is the new line at the end of pg_hba.con 2015-05-22 11:23:11 +02:00
Feike Steenbergen ebfa29b89b Remove "testing" code from the helpers 2015-05-21 14:58:24 +02:00
Feike Steenbergen 836a681157 Fixed the tests, by testing also found an squashed a bug in the StatusPage on python 3. 2015-05-21 14:40:07 +02:00
Alexander Kukushkin 1fb7b2cbe1 Compatibility with python3 2015-05-20 18:19:23 +02:00
Alexander Kukushkin 5d9a3d46a2 Format according to pep8 2015-05-20 18:18:45 +02:00
Oleksii Kliukin becce1dcf0 Merge branch 'master' of https://github.com/zalando/governor into s3_backups 2015-05-20 17:09:39 +02:00
Oleksii Kliukin 363b3c7122 Merge pull request #11 from CyberDem0n/master 
some simple tests
 2015-05-20 16:55:27 +02:00
Oleksii Kliukin 7dd1e295ca re-add the encoding parameter to initdb. 2015-05-20 16:31:26 +02:00
Alexander Kukushkin 3533b790cb Update list of success return codes for PUT and DELETE 2015-05-20 12:03:12 +02:00
Oleksii Kliukin 1c1a15908c use governor yaml parameters and not environment variables to configure backups with WAL-E. 2015-05-20 11:47:20 +02:00
Oleksii Kliukin 8a36f17d21 retry with pg_basebackup if one cannot get data from WAL-E. 2015-05-20 11:47:20 +02:00
Oleksii Kliukin 3898610ea3 Put back the S3/pg_basebackup condition after testing. 2015-05-20 11:47:20 +02:00
Oleksii Kliukin 4af752d668 revert the S3/pg_basebackup choice to unconditionally prefer S3 for testing purposes. 2015-05-20 11:47:20 +02:00
Oleksii Kliukin 2efa97334b returned the choice between S3 and pg_basebackup based on diff_in_bytes 2015-05-20 11:47:20 +02:00
Oleksii Kliukin e02eec91bf save/restore both postgresql.conf and pg_hba.conf for WAL-e backup. 2015-05-20 11:47:20 +02:00