Originally Exhibitor was supported in the ZooKeeper class and
configuration for Exhibitor was taken also from `zookeeper` section in
the yaml config file. In fact, Exhibitor just extends ZooKeeper and now
it is reflected in the code and also Exhibitor got it's own section in
the config.yaml file. It will make it easier to configure Exhibitor
hosts and port via environment variables when PR#211 will be merged.
And some other tricks with overriding handle_one_request and finish
methods from the parent class which were necessary only to make OPTIONS
request from haproxy work with python2, but in fact it was still not
working with python3. Instead of doing all the magic we should simply
give to haproxy what it wants to get: HTTP response code and nothing
more.
It should not call `recover` but `state_handler.follow` directly
(like it is already done in the `clone`).
Otherwise it tries to create a new thread and all async functionality
became borken...
where it is not necessary (test_ha, test_ctl, etc...)
It will simplyfy further refactoring and make it possible to install
implementations of AbstractDCS independant of each other.
Originally we were passing postgresql options as an argument of `pg_ctl
start`. It was nice and convenient because doesn't require to touch
configuration files but this method has one significant drawback: it
wasn't possible to change values of options which were passed as an
arguments without restart (event for the case when option reqires only
reload). Instead of doing that (passing options as arguments) we will:
1) rename original postgresql.conf to postgresql-base.conf
2) write options into postgresql.conf which has `include
'postgresql-base.conf'` on the the third line after comment that this
file is generated by Patroni and you should not change it manually
3) listen_addresses and port are still passed as an arguments to the
pg_ctl (just to be foolproof against ALTER SYSTEM set port to 'random')
In addition to that this commit makes some attributes of `Postgresql`
class private (prefixes them with _)
Somehow when you import only urllib3 it's not possible work with
urllib3.exceptions.HTTPError exception (it looks like it is imported
from some other place. from urllib3.exceptions import HTTPError solves
the problem.
All responses to the client sent by single method `_write_response`
encode('utf-8') is done only inside this method. It make easier to
support existing code because eliminates need to put b'' everywhere
Mostly this tag is necessary to give a hint to load balancer
auto-configuration tool that node should not be included into
LB configuration.
In addition to that Patroni also should not return status_code=200
for a health check if the tag is present and value is not `False`.
Until now get_dcs method was doing absolutely the same job as method
from Patroni class.
In addition to that I did small refactoring and clean up of unit tests
It could take up to 10 seconds to create replication slot.
In addition to that when replica fails to connect to the master via
streaming replication it doesn't retry immediately, but with some
timeout (5 seconds). 10 + 5 == 15 what causes replication check
scenarios fail.
It will capture output to stdout and stderr and print it when test is
failed. Please set LOGLEVEL env variable to INFO or DEBUG if you want
to see everything (as it was before).
Without sudo travis is executing build tasks using docker and waiting
time in this case is really small, usually not longer then 10 seconds.
postgresql-9.5 is installed via addons.apt.packages (without sudo)
But ports 5432 and 5433 are busy. So I had to ajust environment.py to
assign port from higher diapason.
And a few words about build tasks:
First task is used for executing unit tests for all different python versions
The second one is used for executing acceptance tests against etcd
The third one is used for executing acceptance tests against zookeeper
acceptance tests are executed with python2.7 and python3.5
In addition that I've introduced caching of python virtual environment.
It really helps to reduce time needed to install python modules.
According to https://github.com/zalando/patroni/issues/48 'clonefrom'
tag should be boolean and it should be used to mark node as a suitable
for creation of a new replica from. If there are more then one such node
in the cluster (with tag clonefrom=true), one of them will be chosed
randomly.
It was shutdown correctly and I expected such 'join' working, but it was
not, because new leader didn't had enough time to catch up with the
master before promote.
Instead of starting it up in "read-only", it wil wait 2 seconds, to give
a time to somebody to prompte and after it will execute normal `recover`
procedure.
Such situation is causing patroni crash. Usually it was happening during
manual failover, after former master has demoted and `reset_cluster`
method has been called. In this case `fetch_cluster` was `False` and
`_load_cluster` method was returning value from `self._cluster`, which
was `None`.
In addition to that rename confusing `Etcd.client` and
`ZooKeeper.client` into `_client`. This attribute is available from
AbstractDCS and people had wrong impression that it provides the same
interface for different DCS implementations, which is obviously not the
case. For Etcd it has type etcd.Client and for ZooKeeper - KazooClient.
Alexander Kukushkin