In case if one member of a cluster is not available it will retry with
another one and fetch the new cluster configuration. Default timeout for
all requests to etcd is 5 seconds.
Initial cluster configuration can be resolved through:
1) /v2/members call on one of the cluster members on a client port
2) when it is possible to resolve hostname into multiple ip's it will
iterate through list and try to perform action from 1)
3) If there is discovery_srv defined in etcd section of config file it
will resolve peer addresses of all cluster members and will fetch
cluster configuration with using peer protocol by doing /members call on
a peer port
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.
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.
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.
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
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
Alexander Kukushkin