Oleksii Kliukin
bad37a5a21
Always check that cluster is configured correctly right before running pg_rewind.
2015-10-02 10:57:49 +02:00
Oleksii Kliukin
ea910a8987
Make sure pgpass file name is also passed in the PGPASSFILE environment variable.
2015-09-30 18:02:07 +02:00
Oleksii Kliukin
b223319183
use the PATH to get the python interpreter path for the scripts.
2015-09-30 18:00:28 +02:00
Oleksii Kliukin
d7403f999d
Merge branch 'feature/pg_rewind' of https://github.com/zalando/patroni into feature/pg_rewind
2015-09-30 17:13:16 +02:00
Oleksii Kliukin
316e4f8d74
Merge branch 'master' into feature/pg_rewind
2015-09-30 17:12:03 +02:00
Oleksii Kliukin
b4a2ed874f
Merge branch 'master' of https://github.com/zalando/patroni
2015-09-30 16:47:07 +02:00
Oleksii Kliukin
182a90973d
Make sure tests work with the new pytest.
2015-09-30 16:46:53 +02:00
Alexander Kukushkin
2fdf02e0be
Merge pull request #55 from pgexperts/scripts/init-scripts
...
Added Upstart job for running patroni under Ubuntu 14.04 or 12.04.
2015-09-30 16:39:07 +02:00
Oleksii Kliukin
a500781b6d
Mock remove_data_directory in the pg_rewind test.
2015-09-30 16:32:56 +02:00
Josh Berkus
c4cc0be4fe
Added Upstart job for running patroni under Ubuntu 14.04 or 12.04.
2015-09-29 17:35:45 -07:00
Oleksii Kliukin
e39d318732
Eliminate os.system call.
2015-09-28 12:04:06 +02:00
Oleksii Kliukin
d44a54628a
remove the data directory on an unsuccessfull rewind attempt.
2015-09-25 16:00:24 +02:00
Oleksii Kliukin
b62d72a130
Merge branch 'master' into feature/pg_rewind
2015-09-25 13:08:42 +02:00
Oleksii Kliukin
d6c8df45e1
Write the pg_rewind password in pgpass instead of passing it in the command line.
2015-09-25 13:08:12 +02:00
Oleksii Kliukin
36101974cf
Merge pull request #49 from zalando/feature/bootstrap-from-main-loop
...
Run initial cluster bootstrap from the main loop
2015-09-24 15:04:52 +02:00
Oleksii Kliukin
027bcd39ce
Move pg_rewind call into a separate sub.
...
Add a Postgresql method to call pg_rewind.
Improve the test coverage.
2015-09-24 12:46:36 +02:00
Oleksii Kliukin
c8108f221e
Check the exit code of the postgres start when determining whether to run the on_role_change callback.
2015-09-24 11:34:28 +02:00
Oleksii Kliukin
793325cb60
add support for pg_rewind.
2015-09-23 18:38:17 +02:00
Alexander Kukushkin
a4266be3da
remove unused function
2015-09-23 10:59:55 +02:00
Alexander Kukushkin
e83651b57b
Run initial cluster bootstrap from the main loop
2015-09-23 10:55:38 +02:00
Alexander Kukushkin
83c5416c82
Merge pull request #45 from zalando/feature/query-retry
...
Refactor Postgresql.query method to use common retry mechanism
2015-09-22 16:53:31 +02:00
Alexander Kukushkin
d8e51f586f
Merge branch 'feature/query-retry' of github.com:zalando/patroni into feature/query-retry
2015-09-20 13:58:12 +02:00
Alexander Kukushkin
d8982e1e5a
Refactor Postgresql.query method to use common retry mechanism
...
query method in an api.py also needs retry in some cases (for example
when we are running is_healthiest_node check).
In all cases we should retry only when connection is closed or broken.
BUT, the connection status must be checked via cursor.connection (old
implementation was using general connection object for that). For
multi-threaded applications this is not appropriate, because some other
thread might restore connection.
In addition to that I've changed most of the unit tests to use `Mock` and
`patch` where it is possible.
2015-09-20 13:54:30 +02:00
Alexander Kukushkin
9f9cb6005d
Refactor Postgresql.query method to use commont retry mechanism
...
query method in an api.py also needs retry in some cases (for example
when we are running is_healthiest_node check).
In all cases We will retry only when connection is closed or broken, BUT
connection status must be checked via cursor.connection (old
implementation was using general connection object for that). For
multi-threaded applications this is not appropriate, because some other
thread might restore connection.
In appdition to that I've changed most of the unit tests to use `Mock` and
`patch` where it is possible.
2015-09-20 13:41:57 +02:00
Oleksii Kliukin
75be235d49
Merge pull request #40 from zalando/feature/ignore-second-sigterm
...
During shutdown sigterm is sent to all processes every second
2015-09-18 17:06:02 +02:00
Alexander Kukushkin
0a3a6d72cf
Fix etcd unit tests
2015-09-18 15:33:58 +02:00
Alexander Kukushkin
9dfb562a4f
wait for event on leader key a little bit longer than timeout
2015-09-18 15:29:50 +02:00
Alexander Kukushkin
3145e94797
check ignore_sigterm flag instead of setting new handler for SIGTERM
2015-09-18 15:27:09 +02:00
Alexander Kukushkin
afe1a2c92c
During shutdown sigterm is send to all processes every second
...
First time is triggers some shutdown mechanisms in a patroni, but the
second one just terminates patroni. Solution is pretty simple, we need
to ignore all other sigterms which are coming after the first one.
2015-09-18 12:41:36 +02:00
Alexander Kukushkin
8a0af6985f
Merge pull request #35 from zalando/feature/basic-auth
...
Basic-auth and SSL support
2015-09-17 16:22:23 +02:00
Alexander Kukushkin
918ac6a87d
Merge pull request #39 from zalando/feature/reschedule-next-run
...
Feature/reschedule next run of ha cycle depending on return value of `watch`
2015-09-17 16:22:12 +02:00
Alexander Kukushkin
b9ccd411ed
Merge pull request #19 from zalando/feature/faster_shutdown
...
Feature/faster shutdown
2015-09-17 16:21:59 +02:00
Alexander Kukushkin
6530e1f7aa
Remove unused parameter in a is_leader method
2015-09-17 16:11:08 +02:00
Alexander Kukushkin
0b753d25e1
Get rid from is_promoted flag. use role == 'master' instead
2015-09-17 13:57:29 +02:00
Alexander Kukushkin
dbcc5aff9b
Track postgresql role in a Postgresql class
2015-09-16 16:22:10 +02:00
Alexander Kukushkin
e204945b3d
Merge branch 'master' of github.com:zalando/patroni into feature/faster_shutdown
2015-09-16 15:20:53 +02:00
Alexander Kukushkin
0435e36cad
self.cluster = None if unexpected exception occured
2015-09-16 15:14:02 +02:00
Alexander Kukushkin
246f0dbbaf
watch should return true if you are not leader
2015-09-16 14:23:47 +02:00
Alexander Kukushkin
c240fd1ee8
More documentation for watch method
2015-09-16 11:09:53 +02:00
Alexander Kukushkin
7f8e95b334
Next run of ha cycle is rescheduled depending on return value of watch
...
Current etcd implementation does not yet support timeout option when
`wait=true`: https://github.com/coreos/etcd/issues/2468
Originaly I've implemented `watch` method for `Etcd` class in a
following manner: if the leader key was updated just because master
needs to update ttl and watch timeout is not yet expired, I was
recalculating timeout and starting `watch` call once again.
Usually after "restart" we were getting urllib3.exceptions.TimeoutError.
The only possible way to recover after such exception - close socket and
establish a new connection. With pure http it's relatively cheap, but
with https and some kind of authorization on etcd side it would became
rather expensive and should be avoided.
2015-09-16 10:38:34 +02:00
Alexander Kukushkin
01efcc8638
Merge branch 'master' of github.com:zalando/patroni into feature/basic-auth
...
Conflicts:
tests/test_patroni.py
2015-09-15 09:33:18 +02:00
Alexander Kukushkin
a8305079c3
Merge pull request #38 from zalando/bugfix/zookeeper
...
Bugfix/zookeeper
2015-09-15 09:29:18 +02:00
Alexander Kukushkin
90cfcf0c14
Make zookeeper module compatible with python3
2015-09-14 17:14:39 +02:00
Alexander Kukushkin
44a20f12a4
version field is znode is just version, not mzxid
2015-09-14 16:32:45 +02:00
Oleksii Kliukin
bf5737614d
Merge pull request #30 from zalando/feature/cleanup_on_failed_initialization
...
Make sure initialize flag is reset on failure.
2015-09-14 12:57:39 +02:00
Oleksii Kliukin
d69403ab6f
Merge pull request #37 from zalando/feature/zookeeper-fetch-initialize
...
Build Cluster object for ZooKeeper the same way as for Etcd
2015-09-14 12:55:10 +02:00
Oleksii Kliukin
51eacc5042
Handle the case when initialize flag is not set and leader is present.
2015-09-14 12:36:28 +02:00
Alexander Kukushkin
98488a00a2
Remove unused import of KazooException
2015-09-14 12:00:24 +02:00
Alexander Kukushkin
4a081bcb71
Run cancel_initialization with retry
2015-09-14 11:58:10 +02:00
Alexander Kukushkin
209c985420
get_node and get_children should catch only NoNodeError exception.
...
All other exceptions are needed to have retry functionality working
correctly.
2015-09-14 11:45:00 +02:00
Oleksii Kliukin