Release 1.5.2 (#875)

* Update release notes
* Bump version

docs/SETTINGS.rst

@@ -33,6 +33,11 @@ Bootstrap configuration
         -  **restore\_command**: command to restore WAL records from the remote master to standby leader, can be different from the list defined in :ref:`postgresql_settings`
         -  **archive\_cleanup\_command**: cleanup command for standby leader
         -  **recovery\_min\_apply\_delay**: how long to wait before actually apply WAL records on a standby leader
+    -  **slots**: define permanent replication slots. These slots will be preserved during switchover/failover. Patroni will try to create slots before opening connections to the cluster.
+        -  **my_slot_name**: the name of replication slot. It is the responsibility of the operator to make sure that there are no clashes in names between replication slots automatically created by Patroni for members and permanent replication slots.
+            -  **type**: slot type. Could be ``physical`` or ``logical``. If the slot is logical, you have to additionally define ``database`` and ``plugin``.
+               **database**: the database name where logical slots should be created.
+               **plugin**: the plugin name for the logical slot.
 -  **method**: custom script to use for bootstrapping this cluster.
    See :ref:`custom bootstrap methods documentation <custom_bootstrap>` for details.
    When ``initdb`` is specified revert to the default ``initdb`` command. ``initdb`` is also triggered when no ``method``
@@ -69,6 +74,8 @@ Most of the parameters are optional, but you have to specify one of the **host**
 -  **key**: (optional) file with the client key. Can be empty if the key is part of **cert**.
 -  **dc**: (optional) Datacenter to communicate with. By default the datacenter of the host is used.
 -  **checks**: (optional) list of Consul health checks used for the session. If not specified Consul will use "serfHealth" in additional to the TTL based check created by Patroni. Additional checks, in particular the "serfHealth", may cause the leader lock to expire faster than in `ttl` seconds when the leader instance becomes unavailable
+-  **register\_service**: (optional) whether or not to register a service with the name defined by the scope parameter and the tag master, replica or standby-leader depending on the node's role. Defaults to **false**
+-  **service\_check\_interval**: (optional) how often to perform health check against registered url
 
 Etcd
 ----
@@ -144,8 +151,10 @@ PostgreSQL
 
 REST API
 -------- 
--  **connect\_address**: IP address (or hostname) and port, to access the Patroni's REST API. It can serve as a endpoint for HTTP health checks (read below about the "listen" REST API parameter), and also for user queries (either directly or via the REST API), as well as for the health checks done by the cluster members during leader elections (for example, to determine whether the master is still running, or if there is a node which has a WAL position that is ahead of the one doing the query; etc.) The connect_address is put in the member key in DCS, making it possible to translate the member name into the address to connect to its REST API.
--  **listen**: IP address (or hostname) and port that Patroni will listen to for the REST API - to provide also the same health checks and cluster messaging between the participating nodes, as described above. to provide health-check information for HAProxy (or any other load balancer capable of doing a HTTP "OPTION" or "GET" checks)  
+-  **connect\_address**: IP address (or hostname) and port, to access the Patroni's REST API. All the members of the cluster must be able to connect to this address, so unless the Patroni setup is intended for a demo inside the localhost, this address must be a non "localhost" or loopback addres (ie: "localhost" or "127.0.0.1"). It can serve as a endpoint for HTTP health checks (read below about the "listen" REST API parameter), and also for user queries (either directly or via the REST API), as well as for the health checks done by the cluster members during leader elections (for example, to determine whether the master is still running, or if there is a node which has a WAL position that is ahead of the one doing the query; etc.) The connect_address is put in the member key in DCS, making it possible to translate the member name into the address to connect to its REST API.
+
+-  **listen**: IP address (or hostname) and port that Patroni will listen to for the REST API - to provide also the same health checks and cluster messaging between the participating nodes, as described above. to provide health-check information for HAProxy (or any other load balancer capable of doing a HTTP "OPTION" or "GET" checks).
+
 -  **Optional**:
         -  **authentication**:
             -  **username**: Basic-auth username to protect unsafe REST API endpoints.

docs/releases.rst

@@ -3,19 +3,57 @@
 Release notes
 =============
 
-Version 1.5.0                                                                                                                           
--------------                                                                                                                           
-                                                                                                                                        
+Version 1.5.2
+-------------
+
+Compatibility and bugfix release.
+
+- Compatibility with kazoo-2.6.0 (Alexander Kukushkin)
+
+  In order to make sure that requests are performed with an appropriate timeout, Patroni redefines create_connection method from python-kazoo module. The last release of kazoo slightly changed the way how create_connection method is called.
+
+- Fix Patroni crash when Consul cluster loses the leader (Alexander)
+
+  The crash was happening due to incorrect implementation of touch_member method, it should return boolean and not raise any exceptions.
+
+Version 1.5.1
+-------------
+
+This version implements support of permanent replication slots, adds support of pgBackRest and fixes number of bugs.
+
+**New features**
+
+- Permanent replication slots (Alexander Kukushkin)
+
+  Permanent replication slots are preserved on failover/switchover, that is, Patroni on the new primary will create configured replication slots right after doing promote. Slots could be configured with the help of `patronictl edit-config`. The initial configuration could be also done in the :ref:`bootstrap.dcs <settings>`.
+
+- Add pgbackrest support (Yogesh Sharma)
+
+  pgBackrest can restore in existing $PGDATA folder, this allows speedy restore as files which have not changed since last backup are skipped, to support this feature new parameter `keep_data` has been introduced. See :ref:`replica creation method <custom_replica_creation>` section for additional examples.
+
+**Bug fixes**
+
+- A few bugfixes in the "standby cluster" workflow (Alexander)
+
+  Please see https://github.com/zalando/patroni/pull/823 for more details.
+
+- Fix REST API health check when cluster management is paused and DCS is not accessible (Alexander)
+
+  Regression was introduced in https://github.com/zalando/patroni/commit/90cf930036a9d5249265af15d2b787ec7517cf57
+
+Version 1.5.0
+-------------
+
 This version enables Patroni HA cluster to operate in a standby mode, introduces experimental support for running on Windows, and provides a new configuration parameter to register PostgreSQL service in Consul.
 
-**New features**                                                                                                                        
+**New features**
 
-- Standby cluster (Dmitry Dolgov)                                                                                                       
+- Standby cluster (Dmitry Dolgov)
 
   One or more Patroni nodes can form a standby cluster that runs alongside the primary one (i.e. in another datacenter) and consists of standby nodes that replicate from the master in the primary cluster. All PostgreSQL nodes in the standby cluster are replicas; one of those replicas elects itself to replicate directly from the remote master, while the others replicate from it in a cascading manner. More detailed description of this feature and some configuration examples can be found at :ref:`here <standby_cluster>`.
 
 - Register Services in Consul (Pavel Kirillov, Alexander Kukushkin)
-                                                                                                                                        
+
   If `register_service` parameter in the consul :ref:`configuration <consul_settings>` is enabled, the node will register a service with the name `scope` and the tag `master`, `replica` or `standby-leader`.
 
 - Experimental Windows support (Pavel Golub)
@@ -299,8 +337,8 @@ In addition to using Endpoints, Patroni supports ConfigMaps. You can find more i
 
 - Minimize the amount of SELECT's issued by Patroni on every loop of HA cylce (Alexander Kukushkin)
 
-  On every iteration of HA loop Patroni needs to know recovery status and absolute wal position. From now on Patroni will run only single SELECT to get this information instead of two on the replica and three on the master. 
-  
+  On every iteration of HA loop Patroni needs to know recovery status and absolute wal position. From now on Patroni will run only single SELECT to get this information instead of two on the replica and three on the master.
+
 - Remove leader key on shutdown only when we have the lock (Ants)
 
   Unconditional removal was generating unnecessary and missleading exceptions.
@@ -310,7 +348,7 @@ In addition to using Endpoints, Patroni supports ConfigMaps. You can find more i
 - Add version command to patronictl (Ants)
 
   It will show the version of installed Patroni and versions of running Patroni instances (if the cluster name is specified).
-  
+
 - Make optional specifying cluster_name argument for some of patronictl commands (Alexander, Ants)
 
   It will work if patronictl is using usual Patroni configuration file with the ``scope`` defined.
@@ -318,11 +356,11 @@ In addition to using Endpoints, Patroni supports ConfigMaps. You can find more i
 - Show information about scheduled switchover and maintenance mode (Alexander)
 
   Before that it was possible to get this information only from Patroni logs or directly from DCS.
-  
+
 - Improve ``patronictl reinit`` (Alexander)
 
   Sometimes ``patronictl reinit`` refused to proceed when Patroni was busy with other actions, namely trying to start postgres. `patronictl` didn't provide any commands to cancel such long running actions and the only (dangerous) workarond was removing a data directory manually. The new implementation of `reinit` forcefully cancells other long-running actions before proceeding with reinit.
-  
+
 - Implement ``--wait`` flag in ``patronictl pause`` and ``patronictl resume`` (Alexander)
 
   It will make ``patronictl`` wait until the requested action is acknowledged by all nodes in the cluster.
@@ -369,7 +407,7 @@ Version 1.3.6
   After a crash that doesn't clean up postmaster.pid there could be a new process with the same pid, resulting in a false positive for is_running(), which will lead to all kinds of bad behavior.
 
 - Shutdown postgresql before bootstrap when we lost data directory (ainlolcat)
-  
+
   When data directory on the master is forcefully removed, postgres process can still stay alive for some time and prevent the replica created in place of that former master from starting or replicating.
   The fix makes Patroni cache the postmaster pid and its start time and let it terminate the old postmaster in case it is still running after the corresponding data directory has been removed.
 
@@ -439,7 +477,7 @@ Version 1.3.4
 - Pass the consul token as a header (Andrew Colin Kissa)
 
   Headers are now the prefered way to pass the token to the consul `API <https://www.consul.io/api/index.html#authentication>`__.
-  
+
 
 - Advanced configuration for Consul (Alexander Kukushkin)
 

docs/replica_bootstrap.rst

@@ -65,6 +65,19 @@ for outdated backup files. Some people prefer other backup solutions, such as ``
 others, or simply roll their own scripts. In order to accommodate all those use-cases Patroni supports running custom
 scripts to clone a new replica. Those are configured in the ``postgresql`` configuration block:
 
+.. code:: YAML
+
+    postgresql:
+        create_replica_methods:
+            - <method name>
+        <method name>:
+            command: <command name>
+            keep_data: True
+            no_params: True
+            no_master: 1
+
+example: wal_e
+
 .. code:: YAML
 
     postgresql:
@@ -79,6 +92,21 @@ scripts to clone a new replica. Those are configured in the ``postgresql`` confi
         basebackup:
             max-rate: '100M'
 
+example: pgbackrest
+
+.. code:: YAML
+
+    postgresql:
+        create_replica_methods:
+            - pgbackrest
+            - basebackup
+        pgbackrest:
+            command: /usr/bin/pgbackrest --stanza=<scope> --delta restore
+            keep_data: True
+            no_params: True
+        basebackup:
+            max-rate: '100M'
+
 
 The ``create_replica_methods`` defines available replica creation methods and the order of executing them. Patroni will
 stop on the first one that returns 0. Each method should define a separate section in the configuration file, listing the command
@@ -99,6 +127,10 @@ A special ``no_master`` parameter, if defined, allows Patroni to call the replic
 running master or replicas. In that case, an empty string will be passed in a connection string. This is useful for
 restoring the formerly running cluster from the binary backup.
 
+A special ``keep_data`` parameter, if defined, will instuct Patroni to  not clean PGDATA folder before calling restore.
+
+A special ``no_params`` parameter, if defined, restricts passing parameters to custom command.
+
 A ``basebackup`` method is a special case: it will be used if
 ``create_replica_methods`` is empty, although it is possible
 to list it explicitly among the ``create_replica_methods`` methods. This method initializes a new replica with the

extras/README.md

@@ -1,11 +1,11 @@
 ### confd
 
-`confd` directory contains haproxy template files for the [confd](https://github.com/kelseyhightower/confd) -- lightweight configuration management tool
+`confd` directory contains haproxy and pgbouncer template files for the [confd](https://github.com/kelseyhightower/confd) -- lightweight configuration management tool
 You need to copy content of `confd` directory into /etcd/confd and run confd service:
 ```bash
 $ confd -prefix=/service/$PATRONI_SCOPE -backend etcd -node $PATRONI_ETCD_URL -interval=10
 ```
-It will periodically update haproxy.cfg with the actual list of Patroni nodes from `etcd` and "reload" haproxy when it is necessary.
+It will periodically update haproxy.cfg and pgbouncer.ini with the actual list of Patroni nodes from `etcd` and "reload" haproxy and pgbouncer.ini when it is necessary.
 
 
 ### startup-scripts

extras/confd/conf.d/pgbouncer.toml

@@ -0,0 +1,12 @@
+[template]
+prefix = "/service/batman"
+owner = "postgres"
+mode = "0644"
+src = "pgbouncer.tmpl"
+dest = "/etc/pgbouncer/pgbouncer.ini"
+
+reload_cmd = "systemctl reload pgbouncer"
+
+keys = [
+    "/members/","/leader"
+]

extras/confd/templates/pgbouncer.tmpl

@@ -0,0 +1,17 @@
+[databases]
+{{with get "/leader"}}{{$leader := .Value}}{{$leadkey := printf "/members/%s" $leader}}{{with get $leadkey}}{{$data := json .Value}}{{$hostport := base (replace (index (split $data.conn_url "/") 2) "@" "/" -1)}}{{ $host := base (index (split $hostport ":") 0)}}{{ $port := base (index (split $hostport ":") 1)}}* = host={{ $host }} port={{ $port }} pool_size=10{{end}}{{end}}
+
+[pgbouncer]
+logfile = /var/log/postgresql/pgbouncer.log
+pidfile = /var/run/postgresql/pgbouncer.pid
+listen_addr = *
+listen_port = 6432
+unix_socket_dir = /var/run/postgresql
+auth_type = trust
+auth_file = /etc/pgbouncer/userlist.txt
+auth_hba_file = /etc/pgbouncer/pg_hba.txt
+admin_users = pgbouncer
+stats_users = pgbouncer
+pool_mode = session
+max_client_conn = 100
+default_pool_size = 20

extras/startup-scripts/patroni.service

@@ -18,8 +18,14 @@ WorkingDirectory=~
 
 # Where to send early-startup messages from the server
 # This is normally controlled by the global default set by systemd
-# StandardOutput=syslog
+#StandardOutput=syslog
 
+# Pre-commands to start watchdog device
+# Uncomment if watchdog is part of your patroni setup
+#ExecStartPre=-/usr/bin/sudo /sbin/modprobe softdog
+#ExecStartPre=-/usr/bin/sudo /bin/chown postgres /dev/watchdog
+
+# Start the patroni process
 ExecStart=/bin/patroni /etc/patroni.yml
 
 # Send HUP to reload from patroni.yml

features/callback.sh

@@ -0,0 +1,5 @@
+#!/bin/bash
+
+[[ "$3" == "master" ]] || exit
+
+PGPASSWORD=zalando psql -h localhost -U postgres -p $1 -w -tAc "SELECT slot_name FROM pg_replication_slots WHERE slot_type = 'logical'" >> data/postgres0/label

features/environment.py

@@ -182,6 +182,9 @@ class PatroniController(AbstractController):
                         dst[k] = v
             recursive_update(config, custom_config)
 
+        if config['postgresql'].get('callbacks', {}).get('on_role_change'):
+            config['postgresql']['callbacks']['on_role_change'] += ' ' + str(self.__PORT)
+
         with open(patroni_config_path, 'w') as f:
             yaml.safe_dump(config, f, default_flow_style=False)
 

features/standby_cluster.feature

@@ -1,11 +1,27 @@
 Feature: standby cluster
+  Scenario: check permanent logical slots are preserved on failover/switchover
+    Given I start postgres1
+    Then postgres1 is a leader after 10 seconds
+    And I sleep for 2 seconds
+    When I issue a PATCH request to http://127.0.0.1:8009/config with {"slots": {"pm_1": {"type": "physical"}}, "postgresql": {"parameters": {"wal_level": "logical"}}}
+    Then I receive a response code 200
+    And Response on GET http://127.0.0.1:8009/config contains slots after 10 seconds
+    When I issue a PATCH request to http://127.0.0.1:8009/config with {"slots": {"test_logical": {"type": "logical", "database": "postgres", "plugin": "test_decoding"}}}
+    Then I receive a response code 200
+    When I start postgres0 with callback configured
+    Then "members/postgres0" key in DCS has state=running after 10 seconds
+    When I shut down postgres1
+    Then postgres0 is a leader after 10 seconds
+    And I sleep for 2 seconds
+    When I issue a GET request to http://127.0.0.1:8008/
+    Then I receive a response code 200
+    And there is a label with "test_logical" in postgres0 data directory
 
   Scenario: check replication of a single table in a standby cluster
-    Given I start postgres0 without slots sync
-    And I create a replication slot postgres1 on postgres0
-    And I start postgres1 in a standby cluster batman1 as a clone of postgres0
+    Given I start postgres1 in a standby cluster batman1 as a clone of postgres0
     Then postgres1 is a leader of batman1 after 10 seconds
-    When I add the table foo to postgres0
+    When I issue a PATCH request to http://127.0.0.1:8009/config with {"ttl": 20, "loop_wait": 2}
+    And I add the table foo to postgres0
     Then table foo is present on postgres1 after 20 seconds
     When I start postgres2 in a cluster batman1
     Then postgres2 role is the replica after 24 seconds
@@ -13,4 +29,5 @@ Feature: standby cluster
 
   Scenario: check failover
     When I kill postgres1
-	Then postgres2 is replicating from postgres0 after 20 seconds
+    And I kill postmaster on postgres1
+    Then postgres2 is replicating from postgres0 after 20 seconds

features/steps/standby_cluster.py

@@ -1,3 +1,4 @@
+import os
 import time
 
 from behave import step
@@ -8,19 +9,13 @@ SELECT * FROM pg_catalog.pg_stat_replication
 WHERE application_name = '{0}'
 """
 
-create_replication_slot_query = """
-SELECT pg_create_physical_replication_slot('{0}')
-"""
 
-
-@step('I start {name:w} without slots sync')
-def start_patroni_without_slots_sync(context, name):
+@step('I start {name:w} with callback configured')
+def start_patroni_with_callbacks(context, name):
     return context.pctl.start(name, custom_config={
-        "bootstrap": {
-            "dcs": {
-                "postgresql": {
-                    "use_slots": False
-                }
+        "postgresql": {
+            "callbacks": {
+                "on_role_change": "features/callback.sh"
             }
         }
     })
@@ -35,6 +30,10 @@ def start_patroni(context, name, cluster_name):
 
 @step('I start {name:w} in a standby cluster {cluster_name:w} as a clone of {name2:w}')
 def start_patroni_stanby_cluster(context, name, cluster_name, name2):
+    ctl = context.pctl._processes.pop(name, None)
+    # we need to remove patroni.dynamic.json in order to "bootstrap" standby cluster with existing PGDATA
+    if ctl:
+        os.unlink(os.path.join(ctl._data_dir, 'patroni.dynamic.json'))
     port = context.pctl._processes[name2]._connkwargs.get('port')
     return context.pctl.start(name, custom_config={
         "scope": cluster_name,
@@ -43,7 +42,7 @@ def start_patroni_stanby_cluster(context, name, cluster_name, name2):
                 "standby_cluster": {
                     "host": "localhost",
                     "port": port,
-                    "primary_slot_name": "postgres1",
+                    "primary_slot_name": "pm_1",
                 }
             }
         }
@@ -67,12 +66,3 @@ def check_replication_status(context, pg_name1, pg_name2, timeout):
         time.sleep(1)
 
     return False
-
-
-@step('I create a replication slot {slot_name:w} on {pg_name:w}')
-def create_replication_slot(context, slot_name, pg_name):
-    return context.pctl.query(
-        pg_name,
-        create_replication_slot_query.format(slot_name),
-        fail_ok=True
-    )

kubernetes/Dockerfile

@@ -1,32 +1,39 @@
-FROM postgres:9.6
+FROM postgres:10
 MAINTAINER Alexander Kukushkin <[email protected]>
 
 RUN export DEBIAN_FRONTEND=noninteractive \
     && echo 'APT::Install-Recommends "0";\nAPT::Install-Suggests "0";' > /etc/apt/apt.conf.d/01norecommend \
     && apt-get update -y \
     && apt-get upgrade -y \
-    && apt-get install -y git curl jq python-psycopg2 python-yaml python-requests python-six python-pysocks \
-        python-dateutil python-pip python-prettytable python-wheel python-psutil python locales \
+    && apt-cache depends patroni | sed -n -e 's/.* Depends: \(python3-.\+\)$/\1/p' \
+            | grep -Ev '^python3-(sphinx|etcd|consul|kazoo|kubernetes)' \
+            | xargs apt-get install -y curl jq locales git python3-pip python3-wheel \
 
     ## Make sure we have a en_US.UTF-8 locale available
     && localedef -i en_US -c -f UTF-8 -A /usr/share/locale/locale.alias en_US.UTF-8 \
 
-    && pip install setuptools pip --upgrade \
-    && pip install 'git+https://github.com/zalando/patroni.git#egg=patroni[kubernetes]' \
+    && pip3 install setuptools \
+    && pip3 install 'git+https://github.com/zalando/patroni.git#egg=patroni[kubernetes]' \
 
-    && mkdir -p /home/postgres \
-    && chown postgres:postgres /home/postgres \
+    && PGHOME=/home/postgres \
+    && mkdir -p $PGHOME \
+    && chown postgres $PGHOME \
+    && sed -i "s|/var/lib/postgresql.*|$PGHOME:/bin/bash|" /etc/passwd \
+
+    # Set permissions for OpenShift
+    && chmod 775 $PGHOME \
+    && chmod 664 /etc/passwd \
 
     # Clean up
-    && apt-get remove -y git python-pip python-setuptools \
+    && apt-get remove -y git python3-pip python3-wheel \
     && apt-get autoremove -y \
     && apt-get clean -y \
     && rm -rf /var/lib/apt/lists/* /root/.cache
 
-ADD entrypoint.sh callback.py /
+ADD entrypoint.sh /
 
 EXPOSE 5432 8008
 ENV LC_ALL=en_US.UTF-8 LANG=en_US.UTF-8
 USER postgres
 WORKDIR /home/postgres
-CMD ["/bin/bash", "/entrypoint.sh"]
+CMD ["/bin/bash", "/entrypoint.sh"]

kubernetes/callback.py

@@ -1,62 +0,0 @@
-#!/usr/bin/env python
-
-import logging
-import os
-import socket
-import sys
-import time
-
-from kubernetes import client as k8s_client, config as k8s_config
-from urllib3.exceptions import HTTPError
-from six.moves.http_client import HTTPException
-
-logger = logging.getLogger(__name__)
-
-
-class CoreV1Api(k8s_client.CoreV1Api):
-
-    def retry(func):
-        def wrapped(*args, **kwargs):
-            count = 0
-            while True:
-                try:
-                    return func(*args, **kwargs)
-                except (HTTPException, HTTPError, socket.error, socket.timeout):
-                    if count >= 10:
-                        raise
-                    logger.info('Throttling API requests...')
-                    time.sleep(2 ** count * 0.5)
-                    count += 1
-        return wrapped
-
-    @retry
-    def patch_namespaced_endpoints(self, *args, **kwargs):
-        return super(CoreV1Api, self).patch_namespaced_endpoints(*args, **kwargs)
-
-
-def patch_master_endpoint(api, namespace, cluster):
-    addresses = [k8s_client.V1EndpointAddress(ip=os.environ['POD_IP'])]
-    ports = [k8s_client.V1EndpointPort(port=5432)]
-    subsets = [k8s_client.V1EndpointSubset(addresses=addresses, ports=ports)]
-    body = k8s_client.V1Endpoints(subsets=subsets)
-    return api.patch_namespaced_endpoints(cluster, namespace, body)
-
-
-def main():
-    logging.basicConfig(format='%(asctime)s %(levelname)s: %(message)s', level=logging.INFO)
-    if len(sys.argv) != 4 or sys.argv[1] not in ('on_start', 'on_stop', 'on_role_change'):
-        sys.exit('Usage: %s <action> <role> <cluster_name>', sys.argv[0])
-
-    action, role, cluster = sys.argv[1:4]
-
-    k8s_config.load_incluster_config()
-    k8s_api = CoreV1Api()
-
-    namespace = os.environ['KUBERNETES_NAMESPACE']
-
-    if role == 'master' and action in ('on_start', 'on_role_change'):
-        patch_master_endpoint(k8s_api, namespace, cluster)
-
-
-if __name__ == '__main__':
-    main()

kubernetes/entrypoint.sh

@@ -1,5 +1,12 @@
 #!/bin/bash
 
+if [[ $UID -ge 10000 ]]; then
+    GID=$(id -g)
+    sed -e "s/^postgres:x:[^:]*:[^:]*:/postgres:x:$UID:$GID:/" /etc/passwd > /tmp/passwd
+    cat /tmp/passwd > /etc/passwd
+    rm /tmp/passwd
+fi
+
 cat > /home/postgres/patroni.yml <<__EOF__
 bootstrap:
   dcs:
@@ -23,14 +30,10 @@ postgresql:
       password: '${PATRONI_SUPERUSER_PASSWORD}'
     replication:
       password: '${PATRONI_REPLICATION_PASSWORD}'
-  callbacks:
-    on_start: /callback.py
-    on_stop: /callback.py
-    on_role_change: /callback.py
 __EOF__
 
 unset PATRONI_SUPERUSER_PASSWORD PATRONI_REPLICATION_PASSWORD
 export KUBERNETES_NAMESPACE=$PATRONI_KUBERNETES_NAMESPACE
 export POD_NAME=$PATRONI_NAME
 
-exec /usr/bin/python /usr/local/bin/patroni /home/postgres/patroni.yml
+exec /usr/bin/python3 /usr/local/bin/patroni /home/postgres/patroni.yml

kubernetes/openshift-example/README.md

@@ -0,0 +1,49 @@
+# Patroni OpenShift Configuration
+Patroni can be run in OpenShift. Based on the kubernetes configuration, the Dockerfile and Entrypoint has been modified to support the dynamic UID/GID configuration that is applied in OpenShift. This can be run under the standard `restricted` SCC. 
+
+# Examples
+
+## Create test project
+
+```
+oc new-project patroni-test
+```
+
+## Build the image
+
+Note: Update the references when merged upstream. 
+Note: If deploying as a template for multiple users, the following commands should be performed in a shared namespace like `openshift`. 
+
+```
+oc import-image postgres:10 --confirm -n openshift
+oc new-build https://github.com/zalando/patroni --context-dir=kubernetes -n openshift
+```
+
+## Deploy the Image 
+Two configuration templates exist in [templates](templates) directory: 
+- Patroni Ephemeral
+- Patroni Persistent
+
+The only difference is whether or not the statefulset requests persistent storage. 
+
+## Create the Template
+Install the template into the `openshift` namespace if this should be shared across projects: 
+
+```
+oc create -f templates/template_patroni_ephemeral.yml -n openshift
+```
+
+Then, from your own project: 
+
+```
+oc new-app patroni-pgsql-ephemeral
+```
+
+Once the pods are running, two configmaps should be available: 
+
+```
+$ oc get configmap
+NAME                DATA      AGE
+patroniocp-config   0         1m
+patroniocp-leader   0         1m
+```

kubernetes/openshift-example/templates/template_patroni_ephemeral.yml

@@ -0,0 +1,287 @@
+apiVersion: v1
+kind: Template
+metadata:
+  name: patroni-pgsql-ephemeral
+  annotations:
+    description: |-
+      Patroni Postgresql database cluster, without persistent storage.
+
+      WARNING: Any data stored will be lost upon pod destruction. Only use this template for testing.
+    iconClass: icon-postgresql
+    openshift.io/display-name: Patroni Postgresql (Ephemeral)
+    openshift.io/long-description: This template deploys a a patroni postgresql HA cluster without persistent storage.
+    tags: postgresql
+objects:
+- apiVersion: v1
+  kind: Service
+  metadata:
+    creationTimestamp: null
+    labels:
+      application: ${APPLICATION_NAME}
+      cluster-name: ${PATRONI_CLUSTER_NAME}
+    name: ${PATRONI_CLUSTER_NAME}
+  spec:
+    ports:
+    - port: 5432
+      protocol: TCP
+      targetPort: 5432
+    sessionAffinity: None
+    type: ClusterIP
+  status:
+    loadBalancer: {}
+- apiVersion: v1
+  kind: Service
+  metadata:
+    creationTimestamp: null
+    labels:
+      application: ${APPLICATION_NAME}
+      cluster-name: ${PATRONI_CLUSTER_NAME}
+    name: ${PATRONI_MASTER_SERVICE_NAME}
+  spec:
+    ports:
+    - port: 5432
+      protocol: TCP
+      targetPort: 5432
+    selector:
+      application: ${APPLICATION_NAME}
+      cluster-name: ${PATRONI_CLUSTER_NAME}
+      role: master
+    sessionAffinity: None
+    type: ClusterIP
+  status:
+    loadBalancer: {}
+- apiVersion: v1
+  kind: Secret
+  metadata:
+    name: ${PATRONI_CLUSTER_NAME}
+    labels:
+      application: ${APPLICATION_NAME}
+      cluster-name: ${PATRONI_CLUSTER_NAME}
+  stringData:
+    superuser-password: ${PATRONI_SUPERUSER_PASSWORD}
+    replication-password: ${PATRONI_REPLICATION_PASSWORD}
+- apiVersion: v1
+  kind: Service
+  metadata:
+    creationTimestamp: null
+    labels:
+      application: ${APPLICATION_NAME}
+      cluster-name: ${PATRONI_CLUSTER_NAME}
+    name: ${PATRONI_REPLICA_SERVICE_NAME}
+  spec:
+    ports:
+    - port: 5432
+      protocol: TCP
+      targetPort: 5432
+    selector:
+      application: ${APPLICATION_NAME}
+      cluster-name: ${PATRONI_CLUSTER_NAME}
+      role: replica
+    sessionAffinity: None
+    type: ClusterIP
+  status:
+    loadBalancer: {}
+- apiVersion: apps/v1
+  kind: StatefulSet
+  metadata:
+    creationTimestamp: null
+    generation: 3
+    labels:
+      application: ${APPLICATION_NAME}
+      cluster-name: ${PATRONI_CLUSTER_NAME}
+    name: ${APPLICATION_NAME}
+  spec:
+    podManagementPolicy: OrderedReady
+    replicas: 3
+    revisionHistoryLimit: 10
+    selector:
+      matchLabels:
+        application: ${APPLICATION_NAME}
+        cluster-name: ${PATRONI_CLUSTER_NAME}
+    serviceName: ${APPLICATION_NAME}
+    template:
+      metadata:
+        creationTimestamp: null
+        labels:
+          application: ${APPLICATION_NAME}
+          cluster-name: ${PATRONI_CLUSTER_NAME}
+      spec:
+        containers:
+        - env:
+          - name: POD_IP
+            valueFrom:
+              fieldRef:
+                apiVersion: v1
+                fieldPath: status.podIP
+          - name: PATRONI_KUBERNETES_NAMESPACE
+            valueFrom:
+              fieldRef:
+                apiVersion: v1
+                fieldPath: metadata.namespace
+          - name: PATRONI_KUBERNETES_LABELS
+            value: '{application: ${APPLICATION_NAME}, cluster-name: ${PATRONI_CLUSTER_NAME}}'
+          - name: PATRONI_SUPERUSER_USERNAME
+            value: ${PATRONI_SUPERUSER_USERNAME}
+          - name: PATRONI_SUPERUSER_PASSWORD
+            valueFrom:
+              secretKeyRef:
+                key: superuser-password
+                name: ${PATRONI_CLUSTER_NAME}
+          - name: PATRONI_REPLICATION_USERNAME
+            value: ${PATRONI_REPLICATION_USERNAME}
+          - name: PATRONI_REPLICATION_PASSWORD
+            valueFrom:
+              secretKeyRef:
+                key: replication-password
+                name: ${PATRONI_CLUSTER_NAME}
+          - name: PATRONI_SCOPE
+            value: ${PATRONI_CLUSTER_NAME}
+          - name: PATRONI_NAME
+            valueFrom:
+              fieldRef:
+                apiVersion: v1
+                fieldPath: metadata.name
+          - name: PATRONI_POSTGRESQL_DATA_DIR
+            value: /home/postgres/pgdata/pgroot/data
+          - name: PATRONI_POSTGRESQL_PGPASS
+            value: /tmp/pgpass
+          - name: PATRONI_POSTGRESQL_LISTEN
+            value: 0.0.0.0:5432
+          - name: PATRONI_RESTAPI_LISTEN
+            value: 0.0.0.0:8008
+          image: docker-registry.default.svc:5000/${NAMESPACE}/patroni:latest
+          imagePullPolicy: IfNotPresent
+          name: ${APPLICATION_NAME}
+          ports:
+          - containerPort: 8008
+            protocol: TCP
+          - containerPort: 5432
+            protocol: TCP
+          resources: {}
+          terminationMessagePath: /dev/termination-log
+          terminationMessagePolicy: File
+          volumeMounts:
+          - mountPath: /home/postgres/pgdata
+            name: pgdata
+        dnsPolicy: ClusterFirst
+        restartPolicy: Always
+        schedulerName: default-scheduler
+        securityContext: {}
+        serviceAccount: ${SERVICE_ACCOUNT}
+        serviceAccountName: ${SERVICE_ACCOUNT}
+        terminationGracePeriodSeconds: 0
+        volumes:
+        - name: pgdata
+          emptyDir: {}
+    updateStrategy:
+      type: OnDelete
+- apiVersion: v1
+  kind: Endpoints
+  metadata:
+    name: ${APPLICATION_NAME}
+    labels:
+      application: ${APPLICATION_NAME}
+      cluster-name: ${PATRONI_CLUSTER_NAME}
+  subsets: []
+- apiVersion: v1
+  kind: ServiceAccount
+  metadata:
+    name: ${SERVICE_ACCOUNT}
+- apiVersion: rbac.authorization.k8s.io/v1
+  kind: Role
+  metadata:
+    name: ${SERVICE_ACCOUNT}
+  rules:
+  - apiGroups:
+    - ""
+    resources:
+    - configmaps
+    verbs:
+    - create
+    - get
+    - list
+    - patch
+    - update
+    - watch
+    # delete is required only for 'patronictl remove'
+    - delete
+  - apiGroups:
+    - ""
+    resources:
+    - endpoints
+    verbs:
+    - get
+    - patch
+    - update
+    # the following three privileges are necessary only when using endpoints
+    - create
+    - list
+    - watch
+    # delete is required only for for 'patronictl remove'
+    - delete
+  - apiGroups:
+    - ""
+    resources:
+    - pods
+    verbs:
+    - get
+    - list
+    - patch
+    - update
+    - watch
+- apiVersion: rbac.authorization.k8s.io/v1
+  kind: RoleBinding
+  metadata:
+    name: ${SERVICE_ACCOUNT}
+  roleRef:
+    apiGroup: rbac.authorization.k8s.io
+    kind: Role
+    name: ${SERVICE_ACCOUNT}
+  subjects:
+  - kind: ServiceAccount
+    name: ${SERVICE_ACCOUNT}
+parameters:
+- description: The name of the application for labelling all artifacts.
+  displayName: Application Name
+  name: APPLICATION_NAME
+  value: patroni-ephemeral
+- description: The name of the patroni-pgsql cluster.
+  displayName: Cluster Name
+  name: PATRONI_CLUSTER_NAME
+  value: patroni-ephemeral
+- description: The name of the OpenShift Service exposed for the patroni-ephemeral-master container.
+  displayName: Master service name.  
+  name: PATRONI_MASTER_SERVICE_NAME
+  value: patroni-ephemeral-master
+- description: The name of the OpenShift Service exposed for the patroni-ephemeral-replica containers.
+  displayName: Replica service name. 
+  name: PATRONI_REPLICA_SERVICE_NAME
+  value: patroni-ephemeral-replica
+- description: Maximum amount of memory the container can use.
+  displayName: Memory Limit
+  name: MEMORY_LIMIT
+  value: 512Mi
+- description: The OpenShift Namespace where the patroni and postgresql ImageStream resides.
+  displayName: ImageStream Namespace
+  name: NAMESPACE
+  value: openshift
+- description: Username of the superuser account for initialization.
+  displayName: Superuser Username
+  name: PATRONI_SUPERUSER_USERNAME
+  value: postgres
+- description: Password of the superuser account for initialization.
+  displayName: Superuser Passsword
+  name: PATRONI_SUPERUSER_PASSWORD
+  value: postgres
+- description: Username of the replication account for initialization.
+  displayName: Replication Username
+  name: PATRONI_REPLICATION_USERNAME
+  value: postgres
+- description: Password of the replication account for initialization.
+  displayName: Repication Passsword
+  name: PATRONI_REPLICATION_PASSWORD
+  value: postgres
+- description: Service account name used for pods and rolebindings to form a cluster in the project. 
+  displayName: Service Account
+  name: SERVICE_ACCOUNT
+  value: patroniocp

kubernetes/openshift-example/templates/template_patroni_persistent.yaml

@@ -0,0 +1,303 @@
+apiVersion: v1
+kind: Template
+metadata:
+  name: patroni-pgsql-persistent
+  annotations:
+    description: |-
+      Patroni Postgresql database cluster, with persistent storage.
+
+      WARNING: Any data stored will be lost upon pod destruction. Only use this template for testing.
+    iconClass: icon-postgresql
+    openshift.io/display-name: Patroni Postgresql (Persistent)
+    openshift.io/long-description: This template deploys a a patroni postgresql HA cluster without persistent storage.
+    tags: postgresql
+objects:
+- apiVersion: v1
+  kind: Service
+  metadata:
+    creationTimestamp: null
+    labels:
+      application: ${APPLICATION_NAME}
+      cluster-name: ${PATRONI_CLUSTER_NAME}
+    name: ${PATRONI_CLUSTER_NAME}
+  spec:
+    ports:
+    - port: 5432
+      protocol: TCP
+      targetPort: 5432
+    sessionAffinity: None
+    type: ClusterIP
+  status:
+    loadBalancer: {}
+- apiVersion: v1
+  kind: Service
+  metadata:
+    creationTimestamp: null
+    labels:
+      application: ${APPLICATION_NAME}
+      cluster-name: ${PATRONI_CLUSTER_NAME}
+    name: ${PATRONI_MASTER_SERVICE_NAME}
+  spec:
+    ports:
+    - port: 5432
+      protocol: TCP
+      targetPort: 5432
+    selector:
+      application: ${APPLICATION_NAME}
+      cluster-name: ${PATRONI_CLUSTER_NAME}
+      role: master
+    sessionAffinity: None
+    type: ClusterIP
+  status:
+    loadBalancer: {}
+- apiVersion: v1
+  kind: Secret
+  metadata:
+    name: ${PATRONI_CLUSTER_NAME}
+    labels:
+      application: ${APPLICATION_NAME}
+      cluster-name: ${PATRONI_CLUSTER_NAME}
+  stringData:
+    superuser-password: ${PATRONI_SUPERUSER_PASSWORD}
+    replication-password: ${PATRONI_REPLICATION_PASSWORD}
+- apiVersion: v1
+  kind: Service
+  metadata:
+    creationTimestamp: null
+    labels:
+      application: ${APPLICATION_NAME}
+      cluster-name: ${PATRONI_CLUSTER_NAME}
+    name: ${PATRONI_REPLICA_SERVICE_NAME}
+  spec:
+    ports:
+    - port: 5432
+      protocol: TCP
+      targetPort: 5432
+    selector:
+      application: ${APPLICATION_NAME}
+      cluster-name: ${PATRONI_CLUSTER_NAME}
+      role: replica
+    sessionAffinity: None
+    type: ClusterIP
+  status:
+    loadBalancer: {}
+- apiVersion: apps/v1
+  kind: StatefulSet
+  metadata:
+    creationTimestamp: null
+    generation: 3
+    labels:
+      application: ${APPLICATION_NAME}
+      cluster-name: ${PATRONI_CLUSTER_NAME}
+    name: ${APPLICATION_NAME}
+  spec:
+    podManagementPolicy: OrderedReady
+    replicas: 3
+    revisionHistoryLimit: 10
+    selector:
+      matchLabels:
+        application: ${APPLICATION_NAME}
+        cluster-name: ${PATRONI_CLUSTER_NAME}
+    serviceName: ${APPLICATION_NAME}
+    template:
+      metadata:
+        creationTimestamp: null
+        labels:
+          application: ${APPLICATION_NAME}
+          cluster-name: ${PATRONI_CLUSTER_NAME}
+      spec:
+        containers:
+        - env:
+          - name: POD_IP
+            valueFrom:
+              fieldRef:
+                apiVersion: v1
+                fieldPath: status.podIP
+          - name: PATRONI_KUBERNETES_NAMESPACE
+            valueFrom:
+              fieldRef:
+                apiVersion: v1
+                fieldPath: metadata.namespace
+          - name: PATRONI_KUBERNETES_LABELS
+            value: '{application: ${APPLICATION_NAME}, cluster-name: ${PATRONI_CLUSTER_NAME}}'
+          - name: PATRONI_SUPERUSER_USERNAME
+            value: ${PATRONI_SUPERUSER_USERNAME}
+          - name: PATRONI_SUPERUSER_PASSWORD
+            valueFrom:
+              secretKeyRef:
+                key: superuser-password
+                name: ${PATRONI_CLUSTER_NAME}
+          - name: PATRONI_REPLICATION_USERNAME
+            value: ${PATRONI_REPLICATION_USERNAME}
+          - name: PATRONI_REPLICATION_PASSWORD
+            valueFrom:
+              secretKeyRef:
+                key: replication-password
+                name: ${PATRONI_CLUSTER_NAME}
+          - name: PATRONI_SCOPE
+            value: ${PATRONI_CLUSTER_NAME}
+          - name: PATRONI_NAME
+            valueFrom:
+              fieldRef:
+                apiVersion: v1
+                fieldPath: metadata.name
+          - name: PATRONI_POSTGRESQL_DATA_DIR
+            value: /home/postgres/pgdata/pgroot/data
+          - name: PATRONI_POSTGRESQL_PGPASS
+            value: /tmp/pgpass
+          - name: PATRONI_POSTGRESQL_LISTEN
+            value: 0.0.0.0:5432
+          - name: PATRONI_RESTAPI_LISTEN
+            value: 0.0.0.0:8008
+          image: docker-registry.default.svc:5000/${NAMESPACE}/patroni:latest
+          imagePullPolicy: IfNotPresent
+          name: ${APPLICATION_NAME}
+          ports:
+          - containerPort: 8008
+            protocol: TCP
+          - containerPort: 5432
+            protocol: TCP
+          resources: {}
+          terminationMessagePath: /dev/termination-log
+          terminationMessagePolicy: File
+          volumeMounts:
+          - mountPath: /home/postgres/pgdata
+            name: ${APPLICATION_NAME}
+        dnsPolicy: ClusterFirst
+        restartPolicy: Always
+        schedulerName: default-scheduler
+        securityContext: {}
+        serviceAccount: ${SERVICE_ACCOUNT}
+        serviceAccountName: ${SERVICE_ACCOUNT}
+        terminationGracePeriodSeconds: 0
+        volumes:
+        - name: ${APPLICATION_NAME}
+          persistentVolumeClaim: 
+            claimName: ${APPLICATION_NAME}
+    volumeClaimTemplates:
+    - metadata:
+        labels:
+          application: ${APPLICATION_NAME}
+        name: ${APPLICATION_NAME}
+      spec:
+        accessModes:
+        - ReadWriteOnce
+        resources:
+          requests:
+            storage: ${PVC_SIZE}
+    updateStrategy:
+      type: OnDelete
+- apiVersion: v1
+  kind: Endpoints
+  metadata:
+    name: ${APPLICATION_NAME}
+    labels:
+      application: ${APPLICATION_NAME}
+      cluster-name: ${PATRONI_CLUSTER_NAME}
+  subsets: []
+- apiVersion: v1
+  kind: ServiceAccount
+  metadata:
+    name: ${SERVICE_ACCOUNT}
+- apiVersion: rbac.authorization.k8s.io/v1
+  kind: Role
+  metadata:
+    name: ${SERVICE_ACCOUNT}
+  rules:
+  - apiGroups:
+    - ""
+    resources:
+    - configmaps
+    verbs:
+    - create
+    - get
+    - list
+    - patch
+    - update
+    - watch
+    # delete is required only for 'patronictl remove'
+    - delete
+  - apiGroups:
+    - ""
+    resources:
+    - endpoints
+    verbs:
+    - get
+    - patch
+    - update
+    # the following three privileges are necessary only when using endpoints
+    - create
+    - list
+    - watch
+    # delete is required only for for 'patronictl remove'
+    - delete
+  - apiGroups:
+    - ""
+    resources:
+    - pods
+    verbs:
+    - get
+    - list
+    - patch
+    - update
+    - watch
+- apiVersion: rbac.authorization.k8s.io/v1
+  kind: RoleBinding
+  metadata:
+    name: ${SERVICE_ACCOUNT}
+  roleRef:
+    apiGroup: rbac.authorization.k8s.io
+    kind: Role
+    name: ${SERVICE_ACCOUNT}
+  subjects:
+  - kind: ServiceAccount
+    name: ${SERVICE_ACCOUNT}
+parameters:
+- description: The name of the application for labelling all artifacts.
+  displayName: Application Name
+  name: APPLICATION_NAME
+  value: patroni-persistent
+- description: The name of the patroni-pgsql cluster.
+  displayName: Cluster Name
+  name: PATRONI_CLUSTER_NAME
+  value: patroni-persistent
+- description: The name of the OpenShift Service exposed for the patroni-persistent-master container.
+  displayName: Master service name.  
+  name: PATRONI_MASTER_SERVICE_NAME
+  value: patroni-persistent-master
+- description: The name of the OpenShift Service exposed for the patroni-persistent-replica containers.
+  displayName: Replica service name. 
+  name: PATRONI_REPLICA_SERVICE_NAME
+  value: patroni-persistent-replica
+- description: Maximum amount of memory the container can use.
+  displayName: Memory Limit
+  name: MEMORY_LIMIT
+  value: 512Mi
+- description: The OpenShift Namespace where the patroni and postgresql ImageStream resides.
+  displayName: ImageStream Namespace
+  name: NAMESPACE
+  value: openshift
+- description: Username of the superuser account for initialization.
+  displayName: Superuser Username
+  name: PATRONI_SUPERUSER_USERNAME
+  value: postgres
+- description: Password of the superuser account for initialization.
+  displayName: Superuser Passsword
+  name: PATRONI_SUPERUSER_PASSWORD
+  value: postgres
+- description: Username of the replication account for initialization.
+  displayName: Replication Username
+  name: PATRONI_REPLICATION_USERNAME
+  value: postgres
+- description: Password of the replication account for initialization.
+  displayName: Repication Passsword
+  name: PATRONI_REPLICATION_PASSWORD
+  value: postgres
+- description: Service account name used for pods and rolebindings to form a cluster in the project. 
+  displayName: Service Account
+  name: SERVICE_ACCOUNT
+  value: patroni-persistent
+- description: The size of the persistent volume to create. 
+  displayName: Persistent Volume Size
+  name: PVC_SIZE
+  value: 5Gi

kubernetes/openshift-example/test/Jenkinsfile

@@ -0,0 +1,43 @@
+pipeline {
+  agent any
+  stages {
+    stage ('Deploy test pod'){
+      when {
+          expression {
+            openshift.withCluster() {
+              openshift.withProject() {
+                return !openshift.selector( "dc", "pgbench" ).exists()
+              }
+            }
+          }
+        }
+      steps {
+        script {
+            openshift.withCluster() {
+                openshift.withProject() {
+                  def pgbench = openshift.newApp( "https://github.com/stewartshea/docker-pgbench/", "--name=pgbench", "-e PGPASSWORD=postgres", "-e PGUSER=postgres", "-e PGHOST=patroni-persistent-master", "-e PGDATABASE=postgres", "-e TEST_CLIENT_COUNT=20", "-e TEST_DURATION=120" )
+                  def pgbenchdc = openshift.selector( "dc", "pgbench" )
+                  timeout(5) { 
+                    pgbenchdc.rollout().status()
+                    }
+                  }
+                }
+            }
+        }
+    }
+    stage ('Run benchmark Test'){
+      steps {
+          sh '''
+          oc exec $(oc get pods -l app=pgbench | grep Running | awk '{print $1}')  ./test.sh
+          '''
+      }
+    }
+    stage ('Clean up pgtest pod'){
+      steps {
+          sh '''
+          oc delete all -l app=pgbench
+          '''
+      }
+    }
+  }
+}

kubernetes/openshift-example/test/README.md

@@ -0,0 +1,2 @@
+# Jenkins Test
+This pipeline test will create a separate deployment config for a pgbench pod and execute a test against the patroni cluster. This is a sample and should be customized. 

kubernetes/patroni_k8s.yaml

@@ -36,6 +36,8 @@ spec:
           valueFrom:
             fieldRef:
               fieldPath: metadata.namespace
+        - name: PATRONI_KUBERNETES_USE_ENDPOINTS
+          value: 'true'
         - name: PATRONI_KUBERNETES_LABELS
           value: '{application: patroni, cluster-name: patronidemo}'
         - name: PATRONI_SUPERUSER_USERNAME

patroni/api.py

@@ -87,10 +87,14 @@ class RestApiHandler(BaseHTTPRequestHandler):
         replica_status_code = 200 if not patroni.noloadbalance and response.get('role') == 'replica' else 503
         status_code = 503
 
-        if patroni.config.is_standby_cluster and ('standby_leader' in path or 'standby-leader' in path):
+        if patroni.ha.is_standby_cluster() and ('standby_leader' in path or 'standby-leader' in path):
             status_code = 200 if patroni.ha.is_leader() else 503
         elif 'master' in path or 'leader' in path or 'primary' in path:
-            status_code = 200 if patroni.ha.is_leader() else 503
+            # Round-robing across all masters in pause mode if DCS is not accessible
+            if not cluster and patroni.ha.is_paused():
+                status_code = 200 if response['role'] == 'master' else 503
+            else:
+                status_code = 200 if patroni.ha.is_leader() else 503
         elif 'replica' in path:
             status_code = replica_status_code
         elif cluster:  # dcs is available
@@ -408,17 +412,20 @@ class RestApiHandler(BaseHTTPRequestHandler):
                 raise RetryFailedError('')
             stmt = ("WITH replication_info AS ("
                     "SELECT usename, application_name, client_addr, state, sync_state, sync_priority"
-                    " FROM pg_stat_replication) SELECT"
-                    " to_char(pg_postmaster_start_time(), 'YYYY-MM-DD HH24:MI:SS.MS TZ'),"
-                    " CASE WHEN pg_is_in_recovery() THEN 0"
-                    " ELSE ('x' || SUBSTR(pg_{0}file_name(pg_current_{0}_{1}()), 1, 8))::bit(32)::int END,"
-                    " CASE WHEN pg_is_in_recovery() THEN 0"
-                    " ELSE pg_{0}_{1}_diff(pg_current_{0}_{1}(), '0/0')::bigint END,"
-                    " pg_{0}_{1}_diff(COALESCE(pg_last_{0}_receive_{1}(), pg_last_{0}_replay_{1}()), '0/0')::bigint,"
-                    " pg_{0}_{1}_diff(pg_last_{0}_replay_{1}(), '0/0')::bigint,"
-                    " to_char(pg_last_xact_replay_timestamp(), 'YYYY-MM-DD HH24:MI:SS.MS TZ'),"
-                    " pg_is_in_recovery() AND pg_is_{0}_replay_paused(),"
-                    " (SELECT array_to_json(array_agg(row_to_json(ri))) FROM replication_info ri)")
+                    " FROM pg_catalog.pg_stat_replication) SELECT"
+                    " pg_catalog.to_char(pg_catalog.pg_postmaster_start_time(), 'YYYY-MM-DD HH24:MI:SS.MS TZ'),"
+                    " CASE WHEN pg_catalog.pg_is_in_recovery() THEN 0"
+                    " ELSE ('x' || pg_catalog.substr(pg_catalog.pg_{0}file_name("
+                    "pg_catalog.pg_current_{0}_{1}()), 1, 8))::bit(32)::int END,"
+                    " CASE WHEN pg_catalog.pg_is_in_recovery() THEN 0"
+                    " ELSE pg_catalog.pg_{0}_{1}_diff(pg_catalog.pg_current_{0}_{1}(), '0/0')::bigint END,"
+                    " pg_catalog.pg_{0}_{1}_diff(COALESCE(pg_catalog.pg_last_{0}_receive_{1}(),"
+                    " pg_catalog.pg_last_{0}_replay_{1}()), '0/0')::bigint,"
+                    " pg_catalog.pg_{0}_{1}_diff(pg_catalog.pg_last_{0}_replay_{1}(), '0/0')::bigint,"
+                    " pg_catalog.to_char(pg_catalog.pg_last_xact_replay_timestamp(), 'YYYY-MM-DD HH24:MI:SS.MS TZ'),"
+                    " pg_catalog.pg_is_in_recovery() AND pg_catalog.pg_is_{0}_replay_paused(), "
+                    "(SELECT pg_catalog.array_to_json(pg_catalog.array_agg("
+                    "pg_catalog.row_to_json(ri))) FROM replication_info ri)")
 
             row = self.query(stmt.format(self.server.patroni.postgresql.wal_name,
                                          self.server.patroni.postgresql.lsn_name), retry=retry)[0]

patroni/config.py

@@ -9,7 +9,7 @@ import yaml
 
 from collections import defaultdict
 from copy import deepcopy
-from patroni.dcs import ClusterConfig, is_standby_cluster
+from patroni.dcs import ClusterConfig
 from patroni.postgresql import Postgresql
 from patroni.utils import deep_compare, parse_bool, parse_int, patch_config
 from requests.structures import CaseInsensitiveDict
@@ -99,10 +99,6 @@ class Config(object):
     def dynamic_configuration(self):
         return deepcopy(self._dynamic_configuration)
 
-    @property
-    def is_standby_cluster(self):
-        return is_standby_cluster(self._dynamic_configuration.get('standby_cluster'))
-
     def check_mode(self, mode):
         return bool(parse_bool(self._dynamic_configuration.get(mode)))
 

patroni/ctl.py

@@ -252,7 +252,7 @@ def get_cursor(cluster, connect_parameters, role='master', member=None):
     if role == 'any':
         return cursor
 
-    cursor.execute('SELECT pg_is_in_recovery()')
+    cursor.execute('SELECT pg_catalog.pg_is_in_recovery()')
     in_recovery = cursor.fetchone()[0]
 
     if in_recovery and role == 'replica' or not in_recovery and role == 'master':
@@ -392,7 +392,7 @@ def query_member(cluster, cursor, member, role, command, connect_parameters):
             logging.debug(message)
             return [[timestamp(0), message]], None
 
-        cursor.execute('SELECT pg_is_in_recovery()')
+        cursor.execute('SELECT pg_catalog.pg_is_in_recovery()')
         in_recovery = cursor.fetchone()[0]
 
         if in_recovery and role == 'master' or not in_recovery and role == 'replica':

patroni/dcs/__init__.py

@@ -10,16 +10,34 @@ import re
 import six
 import sys
 
-from collections import namedtuple
+from collections import defaultdict, namedtuple
+from copy import deepcopy
 from patroni.exceptions import PatroniException
 from patroni.utils import parse_bool
 from random import randint
 from six.moves.urllib_parse import urlparse, urlunparse, parse_qsl
 from threading import Event, Lock
 
+slot_name_re = re.compile('^[a-z0-9_]{1,63}$')
 logger = logging.getLogger(__name__)
 
 
+def slot_name_from_member_name(member_name):
+    """Translate member name to valid PostgreSQL slot name.
+
+    PostgreSQL replication slot names must be valid PostgreSQL names. This function maps the wider space of
+    member names to valid PostgreSQL names. Names are lowercased, dashes and periods common in hostnames
+    are replaced with underscores, other characters are encoded as their unicode codepoint. Name is truncated
+    to 64 characters. Multiple different member names may map to a single slot name."""
+
+    def replace_char(match):
+        c = match.group(0)
+        return '_' if c in '-.' else "u{:04d}".format(ord(c))
+
+    slot_name = re.sub('[^a-z0-9_]', replace_char, member_name.lower())
+    return slot_name[0:63]
+
+
 def parse_connection_string(value):
     """Original Governor stores connection strings for each cluster members if a following format:
         postgres://{username}:{password}@{connect_address}/postgres
@@ -189,13 +207,12 @@ class RemoteMember(Member):
                 'create_replica_methods',
                 'restore_command',
                 'archive_cleanup_command',
-                'recovery_min_apply_delay')
+                'recovery_min_apply_delay',
+                'no_replication_slot')
 
     def __getattr__(self, name):
-        if name not in RemoteMember.allowed_keys():
-            return
-
-        return self.data.get(name)
+        if name in RemoteMember.allowed_keys():
+            return self.data.get(name)
 
 
 class Leader(namedtuple('Leader', 'index,session,member')):
@@ -275,14 +292,24 @@ class ClusterConfig(namedtuple('ClusterConfig', 'index,data,modify_index')):
     def from_node(index, data, modify_index=None):
         """
         >>> ClusterConfig.from_node(1, '{') is None
-        True
+        False
         """
 
         try:
             data = json.loads(data)
         except (TypeError, ValueError):
-            return None
-        return ClusterConfig(index, data, modify_index or index)
+            data = None
+            modify_index = 0
+        if not isinstance(data, dict):
+            data = {}
+        return ClusterConfig(index, data, index if modify_index is None else modify_index)
+
+    @property
+    def permanent_slots(self):
+        return isinstance(self.data, dict) and (
+                self.data.get('permanent_replication_slots') or
+                self.data.get('permanent_slots') or self.data.get('slots')
+        ) or {}
 
 
 class SyncState(namedtuple('SyncState', 'index,leader,sync_standby')):
@@ -398,8 +425,64 @@ class Cluster(namedtuple('Cluster', 'initialize,config,leader,last_leader_operat
     def is_synchronous_mode(self):
         return self.check_mode('synchronous_mode')
 
-    def is_standby_cluster(self):
-        return is_standby_cluster(self.config and self.config.data.get('standby_cluster'))
+    def get_replication_slots(self, name, role):
+        # if the replicatefrom tag is set on the member - we should not create the replication slot for it on
+        # the current master, because that member would replicate from elsewhere. We still create the slot if
+        # the replicatefrom destination member is currently not a member of the cluster (fallback to the
+        # master), or if replicatefrom destination member happens to be the current master
+        if role in ('master', 'standby_leader'):
+            slot_members = [m.name for m in self.members if m.name != name and
+                            (m.replicatefrom is None or m.replicatefrom == name or
+                             not self.has_member(m.replicatefrom))]
+            permanent_slots = (self.config and self.config.permanent_slots or {}).copy()
+        else:
+            # only manage slots for replicas that replicate from this one, except for the leader among them
+            slot_members = [m.name for m in self.members if m.replicatefrom == name and m.name != self.leader.name]
+            permanent_slots = {}
+
+        slots = {slot_name_from_member_name(name): {'type': 'physical'} for name in slot_members}
+
+        if len(slots) < len(slot_members):
+            # Find which names are conflicting for a nicer error message
+            slot_conflicts = defaultdict(list)
+            for name in slot_members:
+                slot_conflicts[slot_name_from_member_name(name)].append(name)
+            logger.error("Following cluster members share a replication slot name: %s",
+                         "; ".join("{} map to {}".format(", ".join(v), k)
+                                   for k, v in slot_conflicts.items() if len(v) > 1))
+
+        # "merge" replication slots for members with permanent_replication_slots
+        for name, value in permanent_slots.items():
+            if not slot_name_re.match(name):
+                logger.error("Invalid permanent replication slot name '%s'", name)
+                logger.error("Slot name may only contain lower case letters, numbers, and the underscore chars")
+                continue
+
+            if name in slots:
+                logger.error("Permanent replication slot {'%s': %s} is conflicting with" +
+                             " physical replication slot for cluster member", name, value)
+                continue
+
+            value = deepcopy(value)
+            if not value:
+                value = {'type': 'physical'}
+
+            if isinstance(value, dict):
+                if 'type' not in value:
+                    value['type'] = 'logical' if value.get('database') and value.get('plugin') else 'physical'
+
+                if value['type'] == 'physical' or value['type'] == 'logical' \
+                        and value.get('database') and value.get('plugin'):
+                    slots[name] = value
+                    continue
+
+            logger.error("Bad value for slot '%s' in permanent_slots: %s", name, permanent_slots[name])
+
+        return slots
+
+    def has_permanent_logical_slots(self, name):
+        slots = self.get_replication_slots(name, 'master').values()
+        return any(v for v in slots if v.get("type") == "logical")
 
 
 @six.add_metaclass(abc.ABCMeta)
@@ -537,7 +620,7 @@ class AbstractDCS(object):
         You have to use CAS (Compare And Swap) operation in order to update leader key,
         for example for etcd `prevValue` parameter must be used."""
 
-    def update_leader(self, last_operation):
+    def update_leader(self, last_operation, access_is_restricted=False):
         """Update leader key (or session) ttl and optime/leader
 
         :param last_operation: absolute xlog location in bytes
@@ -654,14 +737,3 @@ class AbstractDCS(object):
 
         self.event.wait(timeout)
         return self.event.isSet()
-
-
-def is_standby_cluster(config):
-    """ Check whether or not provided configuration describes a standby cluster.
-        Config can be both patroni config or cluster.config.data
-    """
-    return isinstance(config, dict) and (
-        config.get('host') or
-        config.get('port') or
-        config.get('restore_command')
-    )

patroni/dcs/consul.py

@@ -338,17 +338,15 @@ class Consul(AbstractDCS):
             logger.exception('get_cluster')
             raise ConsulError('Consul is not responding properly')
 
+    @catch_consul_errors
     def touch_member(self, data, ttl=None, permanent=False):
         cluster = self.cluster
         member = cluster and cluster.get_member(self._name, fallback_to_leader=False)
         create_member = not permanent and self.refresh_session()
 
         if member and (create_member or member.session != self._session):
-            try:
-                self._client.kv.delete(self.member_path)
-                create_member = True
-            except Exception:
-                return False
+            self._client.kv.delete(self.member_path)
+            create_member = True
 
         if not create_member and member and deep_compare(data, member.data):
             return True

patroni/dcs/kubernetes.py

@@ -152,7 +152,8 @@ class Kubernetes(AbstractDCS):
 
             # get global dynamic configuration
             config = ClusterConfig.from_node(metadata and metadata.resource_version,
-                                             annotations.get(self._CONFIG) or '{}')
+                                             annotations.get(self._CONFIG) or '{}',
+                                             metadata.resource_version if self._CONFIG in annotations else 0)
 
             # get timeline history
             history = TimelineHistory.from_node(metadata and metadata.resource_version,
@@ -279,7 +280,7 @@ class Kubernetes(AbstractDCS):
     def _update_leader(self):
         """Unused"""
 
-    def update_leader(self, last_operation):
+    def update_leader(self, last_operation, access_is_restricted=False):
         now = datetime.datetime.now(tzutc).isoformat()
         annotations = {self._LEADER: self._name, 'ttl': str(self._ttl), 'renewTime': now,
                        'acquireTime': self._leader_observed_record.get('acquireTime') or now,
@@ -287,7 +288,11 @@ class Kubernetes(AbstractDCS):
         if last_operation:
             annotations[self._OPTIME] = last_operation
 
-        ret = self.patch_or_create(self.leader_path, annotations, self._leader_resource_version, subsets=self.__subsets)
+        subsets = self.__subsets
+        if subsets is not None and access_is_restricted:
+            subsets = []
+
+        ret = self.patch_or_create(self.leader_path, annotations, self._leader_resource_version, subsets=subsets)
         if ret:
             self._leader_resource_version = ret.metadata.resource_version
         return ret
@@ -307,7 +312,8 @@ class Kubernetes(AbstractDCS):
             else:
                 annotations['acquireTime'] = self._leader_observed_record.get('acquireTime') or now
             annotations['transitions'] = str(transitions)
-        ret = self.patch_or_create(self.leader_path, annotations, self._leader_resource_version, subsets=self.__subsets)
+        subsets = [] if self.__subsets else None
+        ret = self.patch_or_create(self.leader_path, annotations, self._leader_resource_version, subsets=subsets)
         if ret:
             self._leader_resource_version = ret.metadata.resource_version
         else:

patroni/dcs/zookeeper.py

@@ -37,7 +37,9 @@ class PatroniSequentialThreadingHandler(SequentialThreadingHandler):
                      `connect_timeout` (negotiated session timeout) as the second element."""
 
         args = list(args)
-        if len(args) == 1:
+        if len(args) == 0:  # kazoo 2.6.0 slightly changed the way how it calls create_connection method
+            kwargs['timeout'] = max(self._connect_timeout, kwargs.get('timeout', self._connect_timeout*10)/10.0)
+        elif len(args) == 1:
             args.append(self._connect_timeout)
         else:
             args[1] = max(self._connect_timeout, args[1]/10.0)

patroni/ha.py

@@ -12,7 +12,7 @@ from collections import namedtuple
 from multiprocessing.pool import ThreadPool
 from patroni.async_executor import AsyncExecutor, CriticalTask
 from patroni.exceptions import DCSError, PostgresConnectionException, PatroniException
-from patroni.postgresql import ACTION_ON_START
+from patroni.postgresql import ACTION_ON_START, ACTION_ON_ROLE_CHANGE
 from patroni.utils import polling_loop, tzutc
 from patroni.dcs import RemoteMember
 from threading import RLock
@@ -61,6 +61,7 @@ class Ha(object):
         self.old_cluster = None
         self._is_leader = False
         self._is_leader_lock = RLock()
+        self._leader_access_is_restricted = False
         self._was_paused = False
         self._leader_timeline = None
         self.recovering = False
@@ -91,14 +92,30 @@ class Ha(object):
     def is_paused(self):
         return self.check_mode('pause')
 
+    def get_standby_cluster_config(self):
+        if self.cluster and self.cluster.config and self.cluster.config.modify_index:
+            config = self.cluster.config.data
+        else:
+            config = self.patroni.config.dynamic_configuration
+        return config.get('standby_cluster')
+
+    def is_standby_cluster(self):
+        config = self.get_standby_cluster_config()
+        # Check whether or not provided configuration describes a standby cluster
+        return isinstance(config, dict) and (config.get('host') or config.get('port') or config.get('restore_command'))
+
     def is_leader(self):
         with self._is_leader_lock:
-            return self._is_leader
+            return self._is_leader and not self._leader_access_is_restricted
 
     def set_is_leader(self, value):
         with self._is_leader_lock:
             self._is_leader = value
 
+    def set_leader_access_is_restricted(self, value):
+        with self._is_leader_lock:
+            self._leader_access_is_restricted = value
+
     def load_cluster_from_dcs(self):
         cluster = self.dcs.get_cluster()
 
@@ -113,6 +130,7 @@ class Ha(object):
         self._leader_timeline = None if cluster.is_unlocked() else cluster.leader.timeline
 
     def acquire_lock(self):
+        self.set_leader_access_is_restricted(self.cluster.has_permanent_logical_slots(self.state_handler.name))
         ret = self.dcs.attempt_to_acquire_leader()
         self.set_is_leader(ret)
         return ret
@@ -124,7 +142,7 @@ class Ha(object):
                 last_operation = self.state_handler.last_operation()
             except Exception:
                 logger.exception('Exception when called state_handler.last_operation()')
-        ret = self.dcs.update_leader(last_operation)
+        ret = self.dcs.update_leader(last_operation, self._leader_access_is_restricted)
         self.set_is_leader(ret)
         if ret:
             self.watchdog.keepalive()
@@ -151,6 +169,10 @@ class Ha(object):
                 'state': self.state_handler.state,
                 'role': self.state_handler.role
             }
+
+            # following two lines are mainly necessary for consul, to avoid creation of master service
+            if data['role'] == 'master' and not self.is_leader():
+                data['role'] = 'promoted'
             tags = self.get_effective_tags()
             if tags:
                 data['tags'] = tags
@@ -202,7 +224,7 @@ class Ha(object):
                 self.state_handler.bootstrapping = True
                 self._post_bootstrap_task = CriticalTask()
 
-                if self.patroni.config.is_standby_cluster:
+                if self.is_standby_cluster():
                     self._async_executor.schedule('bootstrap_standby_leader')
                     self._async_executor.run_async(self.bootstrap_standby_leader)
                     return 'trying to bootstrap a new standby leader'
@@ -228,8 +250,7 @@ class Ha(object):
             not a real master, but a 'standby leader', that will take base backup
             from a remote master and start follow it.
         """
-        patroni_config = self.patroni.config.dynamic_configuration
-        clone_source = self.get_remote_master(patroni_config)
+        clone_source = self.get_remote_master()
         msg = 'clone from remote master {0}'.format(clone_source.conn_url)
         result = self.clone(clone_source, msg)
         self._post_bootstrap_task.complete(result)
@@ -239,9 +260,10 @@ class Ha(object):
         return result
 
     def _handle_rewind(self):
-        if self.state_handler.rewind_needed_and_possible(self.cluster.leader):
-            self._async_executor.schedule('running pg_rewind from ' + self.cluster.leader.name)
-            self._async_executor.run_async(self.state_handler.rewind, (self.cluster.leader,))
+        leader = self.get_remote_master() if self.is_standby_cluster() else self.cluster.leader
+        if self.state_handler.rewind_needed_and_possible(leader):
+            self._async_executor.schedule('running pg_rewind from ' + leader.name)
+            self._async_executor.run_async(self.state_handler.rewind, (leader,))
             return True
 
     def recover(self):
@@ -273,16 +295,24 @@ class Ha(object):
 
         self.load_cluster_from_dcs()
 
-        if self.has_lock():
-            msg = "starting as readonly because i had the session lock"
-            node_to_follow = None
-        else:
+        if self.is_standby_cluster() or not self.has_lock():
             if not self.state_handler.rewind_executed:
                 self.state_handler.trigger_check_diverged_lsn()
             if self._handle_rewind():
                 return self._async_executor.scheduled_action
-            msg = "starting as a secondary"
-            node_to_follow = self._get_node_to_follow(self.cluster)
+
+            if self.has_lock():  # in standby cluster
+                msg = "starting as a standby leader because i had the session lock"
+                node_to_follow = self._get_node_to_follow(self.cluster)
+            elif self.is_standby_cluster() and self.cluster.is_unlocked():
+                msg = "trying to follow a remote master because standby cluster is unhealthy"
+                node_to_follow = self.get_remote_master()
+            else:
+                msg = "starting as a secondary"
+                node_to_follow = self._get_node_to_follow(self.cluster)
+        elif self.has_lock():
+            msg = "starting as readonly because i had the session lock"
+            node_to_follow = None
 
         self.recovering = True
 
@@ -295,8 +325,8 @@ class Ha(object):
         # try to follow the node mentioned there, otherwise, follow the leader.
         is_leader = self.cluster.leader and self.state_handler.name == self.cluster.leader.name
 
-        if self.cluster.is_standby_cluster() and is_leader:
-            node_to_follow = self.get_remote_master(cluster.config.data)
+        if self.is_standby_cluster() and (is_leader or self.cluster.is_unlocked()):
+            node_to_follow = self.get_remote_master()
         elif self.patroni.replicatefrom and self.patroni.replicatefrom != self.state_handler.name:
             node_to_follow = cluster.get_member(self.patroni.replicatefrom)
         else:
@@ -479,8 +509,10 @@ class Ha(object):
                     return 'Postponing promotion because synchronous replication state was updated by somebody else'
                 self.state_handler.set_synchronous_standby('*' if self.is_synchronous_mode_strict() else None)
             if self.state_handler.role != 'master':
+                self.set_leader_access_is_restricted(self.cluster.has_permanent_logical_slots(self.state_handler.name))
                 self._async_executor.schedule('promote')
-                self._async_executor.run_async(self.state_handler.promote, args=(self.dcs.loop_wait,))
+                self._async_executor.run_async(self.state_handler.promote,
+                                               args=(self.dcs.loop_wait, self._leader_access_is_restricted))
             return promote_message
 
     @staticmethod
@@ -780,7 +812,7 @@ class Ha(object):
                         self.dcs.manual_failover('', '')
                 self.load_cluster_from_dcs()
 
-                if self.cluster.is_standby_cluster():
+                if self.is_standby_cluster():
                     # standby leader disappeared, and this is a healthiest
                     # replica, so it should become a new standby leader.
                     # This imply that we need to start following a remote master
@@ -817,12 +849,17 @@ class Ha(object):
                 self.dcs.reset_cluster()
                 return 'removed leader lock because postgres is not running as master'
 
+            if self.state_handler.is_leader() and self._leader_access_is_restricted:
+                self.state_handler.sync_replication_slots(self.cluster)
+                self.state_handler.call_nowait(ACTION_ON_ROLE_CHANGE)
+                self.set_leader_access_is_restricted(False)
+
             if self.update_lock(True):
                 msg = self.process_manual_failover_from_leader()
                 if msg is not None:
                     return msg
 
-                if self.cluster.is_standby_cluster():
+                if self.is_standby_cluster():
                     # in case of standby cluster we don't really need to
                     # enforce anything, since the leader is not a master.
                     # So just remind the role.
@@ -962,7 +999,8 @@ class Ha(object):
 
     def _do_reinitialize(self, cluster):
         self.state_handler.stop('immediate')
-        self.state_handler.remove_data_directory()
+        # Commented redundant data directory cleanup here
+        # self.state_handler.remove_data_directory()
 
         clone_member = self.cluster.get_clone_member(self.state_handler.name)
         member_role = 'leader' if clone_member == self.cluster.leader else 'replica'
@@ -1055,6 +1093,7 @@ class Ha(object):
         if not self.watchdog.activate():
             logger.error('Cancelling bootstrap because watchdog activation failed')
             self.cancel_initialization()
+        self.state_handler.sync_replication_slots(self.cluster)
         self.dcs.take_leader()
         self.set_is_leader(True)
         self.state_handler.call_nowait(ACTION_ON_START)
@@ -1213,11 +1252,11 @@ class Ha(object):
                 # stops PostgreSQL, therefore, we only reload replication slots if no
                 # asynchronous processes are running (should be always the case for the master)
                 if not self._async_executor.busy and not self.state_handler.is_starting():
+                    self.state_handler.sync_replication_slots(self.cluster)
                     if not self.state_handler.cb_called:
                         if not self.state_handler.is_leader():
                             self.state_handler.trigger_check_diverged_lsn()
                         self.state_handler.call_nowait(ACTION_ON_START)
-                    self.state_handler.sync_replication_slots(self.cluster)
         except DCSError:
             dcs_failed = True
             logger.error('Error communicating with DCS')
@@ -1275,15 +1314,14 @@ class Ha(object):
         This usually happens on the master or if the node is running async action"""
         self.dcs.event.set()
 
-    def get_remote_master(self, config):
+    def get_remote_master(self):
         """ In case of standby cluster this will tel us from which remote
             master to stream. Config can be both patroni config or
             cluster.config.data
         """
-        config = config or (self.config is not None and self.config.data)
+        cluster_params = self.get_standby_cluster_config()
 
-        if config and config.get('standby_cluster'):
-            cluster_params = config.get('standby_cluster')
+        if cluster_params:
             unique_name = 'remote_master:{}'.format(uuid.uuid1())
             data = {
                 'conn_kwargs': {

patroni/postgresql.py

@@ -15,7 +15,7 @@ from patroni.callback_executor import CallbackExecutor
 from patroni.exceptions import PostgresConnectionException, PostgresException
 from patroni.utils import compare_values, parse_bool, parse_int, Retry, RetryFailedError, polling_loop, split_host_port
 from patroni.postmaster import PostmasterProcess
-from patroni.dcs import RemoteMember
+from patroni.dcs import slot_name_from_member_name, RemoteMember
 from requests.structures import CaseInsensitiveDict
 from six import string_types
 from six.moves.urllib.parse import quote_plus
@@ -37,14 +37,16 @@ STATE_UNKNOWN = 'unknown'
 
 STOP_POLLING_INTERVAL = 1
 REWIND_STATUS = type('Enum', (), {'INITIAL': 0, 'CHECK': 1, 'NEED': 2, 'NOT_NEED': 3, 'SUCCESS': 4, 'FAILED': 5})
-sync_standby_name_re = re.compile('^[A-Za-z_][A-Za-z_0-9\$]*$')
+sync_standby_name_re = re.compile(r'^[A-Za-z_][A-Za-z_0-9\$]*$')
 
-cluster_info_query = ("SELECT CASE WHEN pg_is_in_recovery() THEN 0 "
-                      "ELSE ('x' || SUBSTR(pg_{0}file_name(pg_current_{0}_{1}()), 1, 8))::bit(32)::int END, "
-                      "CASE WHEN pg_is_in_recovery() THEN GREATEST("
-                      " pg_{0}_{1}_diff(COALESCE(pg_last_{0}_receive_{1}(), '0/0'), '0/0')::bigint,"
-                      " pg_{0}_{1}_diff(pg_last_{0}_replay_{1}(), '0/0')::bigint)"
-                      "ELSE pg_{0}_{1}_diff(pg_current_{0}_{1}(), '0/0')::bigint END")
+cluster_info_query = ("SELECT CASE WHEN pg_catalog.pg_is_in_recovery() THEN 0 "
+                      "ELSE ('x' || pg_catalog.substr(pg_catalog.pg_{0}file_name("
+                      "pg_catalog.pg_current_{0}_{1}()), 1, 8))::bit(32)::int END, "
+                      "CASE WHEN pg_catalog.pg_is_in_recovery() THEN GREATEST("
+                      " pg_catalog.pg_{0}_{1}_diff(COALESCE("
+                      "pg_catalog.pg_last_{0}_receive_{1}(), '0/0'), '0/0')::bigint,"
+                      " pg_catalog.pg_{0}_{1}_diff(pg_catalog.pg_last_{0}_replay_{1}(), '0/0')::bigint)"
+                      "ELSE pg_catalog.pg_{0}_{1}_diff(pg_catalog.pg_current_{0}_{1}(), '0/0')::bigint END")
 
 
 def quote_ident(value):
@@ -52,22 +54,6 @@ def quote_ident(value):
     return value if sync_standby_name_re.match(value) else '"' + value + '"'
 
 
-def slot_name_from_member_name(member_name):
-    """Translate member name to valid PostgreSQL slot name.
-
-    PostgreSQL replication slot names must be valid PostgreSQL names. This function maps the wider space of
-    member names to valid PostgreSQL names. Names are lowercased, dashes and periods common in hostnames
-    are replaced with underscores, other characters are encoded as their unicode codepoint. Name is truncated
-    to 64 characters. Multiple different member names may map to a single slot name."""
-
-    def replace_char(match):
-        c = match.group(0)
-        return '_' if c in '-.' else "u{:04d}".format(ord(c))
-
-    slot_name = re.sub('[^a-z0-9_]', replace_char, member_name.lower())
-    return slot_name[0:63]
-
-
 @contextmanager
 def null_context():
     yield
@@ -154,7 +140,7 @@ class Postgresql(object):
         self._connection = None
         self._cursor_holder = None
         self._sysid = None
-        self._replication_slots = []  # list of already existing replication slots
+        self._replication_slots = {}  # already existing replication slots
         self.retry = Retry(max_tries=-1, deadline=config['retry_timeout']/2.0, max_delay=1,
                            retry_exceptions=PostgresConnectionException)
 
@@ -324,8 +310,8 @@ class Postgresql(object):
                     changes['wal_segment_size'] = '16384kB'
                 # XXX: query can raise an exception
                 for r in self.query("""SELECT name, setting, unit, vartype, context
-                                         FROM pg_settings
-                                        WHERE LOWER(name) IN (""" + ', '.join(['%s'] * len(changes)) + """)
+                                         FROM pg_catalog.pg_settings
+                                        WHERE pg_catalog.lower(name) IN (""" + ', '.join(['%s'] * len(changes)) + """)
                                         ORDER BY 1 DESC""", *(k.lower() for k in changes.keys())):
                     if r[4] == 'internal':
                         if r[0] == 'wal_segment_size':
@@ -706,8 +692,10 @@ class Postgresql(object):
                     # if basebackup succeeds, exit with success
                     break
             else:
-                if not self.data_directory_empty():
+                if not self.data_directory_empty() and not self.config.get(replica_method, {}).get('keep_data', False):
                     self.remove_data_directory()
+                else:
+                    logger.info('Leaving data directory uncleaned')
 
                 cmd = replica_method
                 method_config = {}
@@ -720,10 +708,14 @@ class Postgresql(object):
                     cmd = method_config.pop('command', cmd)
 
                 # add the default parameters
-                method_config.update({"scope": self.scope,
-                                      "role": "replica",
-                                      "datadir": self._data_dir,
-                                      "connstring": connstring})
+                if not method_config.get('no_params', False):
+                    method_config.update({"scope": self.scope,
+                                          "role": "replica",
+                                          "datadir": self._data_dir,
+                                          "connstring": connstring})
+                else:
+                    for param in ('no_params', 'no_master', 'keep_data'):
+                        method_config.pop(param, None)
                 params = ["--{0}={1}".format(arg, val) for arg, val in method_config.items()]
                 try:
                     # call script with the full set of parameters
@@ -951,7 +943,7 @@ class Postgresql(object):
             with self._get_connection_cursor(**connect_kwargs) as cur:
                 cur.execute("SET statement_timeout = 0")
                 if check_not_is_in_recovery:
-                    cur.execute('SELECT pg_is_in_recovery()')
+                    cur.execute('SELECT pg_catalog.pg_is_in_recovery()')
                     if cur.fetchone()[0]:
                         return 'is_in_recovery=true'
                 return cur.execute('CHECKPOINT')
@@ -1204,7 +1196,7 @@ class Postgresql(object):
             ('user', r.get('user')),
             ('host', r.get('host')),
             ('port', r.get('port')),
-            ('dbname', r.get('database')),
+            ('dbname', r.get('database') or self._database),
             ('sslmode', 'prefer'),
             ('sslcompression', '1'),
         ]
@@ -1249,8 +1241,9 @@ class Postgresql(object):
                 yield cur
 
     @contextmanager
-    def _get_replication_connection_cursor(self, host='localhost', port=5432, **kwargs):
-        with self._get_connection_cursor(host=host, port=int(port), database=self._database, replication=1,
+    def _get_replication_connection_cursor(self, host='localhost', port=5432, database=None, **kwargs):
+        database = database or self._database
+        with self._get_connection_cursor(host=host, port=int(port), database=database, replication='database',
                                          user=self._replication['username'], password=self._replication['password'],
                                          connect_timeout=3, options='-c statement_timeout=2000') as cur:
             yield cur
@@ -1258,7 +1251,7 @@ class Postgresql(object):
     def check_leader_is_not_in_recovery(self, **kwargs):
         try:
             with self._get_connection_cursor(connect_timeout=3, options='-c statement_timeout=2000', **kwargs) as cur:
-                cur.execute('SELECT pg_is_in_recovery()')
+                cur.execute('SELECT pg_catalog.pg_is_in_recovery()')
                 if not cur.fetchone()[0]:
                     return True
                 logger.info('Leader is still in_recovery and therefore can\'t be used for rewind')
@@ -1369,10 +1362,10 @@ class Postgresql(object):
         history_path = 'pg_{0}/{1:08X}.history'.format(self.wal_name, timeline)
         try:
             cursor = self._cursor()
-            cursor.execute('SELECT isdir, modification FROM pg_stat_file(%s)', (history_path,))
+            cursor.execute('SELECT isdir, modification FROM pg_catalog.pg_stat_file(%s)', (history_path,))
             isdir, modification = cursor.fetchone()
             if not isdir:
-                cursor.execute('SELECT pg_read_file(%s)', (history_path,))
+                cursor.execute('SELECT pg_catalog.pg_read_file(%s)', (history_path,))
                 history = list(self.parse_history(cursor.fetchone()[0]))
                 if history[-1][0] == timeline - 1:
                     history[-1].append(modification.isoformat())
@@ -1501,7 +1494,7 @@ class Postgresql(object):
             if data.get('Database cluster state') == 'in production':
                 return True
 
-    def promote(self, wait_seconds):
+    def promote(self, wait_seconds, access_is_restricted=False):
         if self.role == 'master':
             return True
         ret = self.pg_ctl('promote', '-W')
@@ -1509,7 +1502,8 @@ class Postgresql(object):
             self.set_role('master')
             logger.info("cleared rewind state after becoming the leader")
             self._rewind_state = REWIND_STATUS.INITIAL
-            self.call_nowait(ACTION_ON_ROLE_CHANGE)
+            if not access_is_restricted:
+                self.call_nowait(ACTION_ON_ROLE_CHANGE)
             ret = self._wait_promote(wait_seconds)
         return ret
 
@@ -1542,61 +1536,88 @@ $$""".format(name, ' '.join(options)), name, password, password)
 
     def load_replication_slots(self):
         if self.use_slots and self._schedule_load_slots:
-            cursor = self._query("SELECT slot_name FROM pg_replication_slots WHERE slot_type='physical'")
-            self._replication_slots = [r[0] for r in cursor]
+            replication_slots = {}
+            cursor = self._query('SELECT slot_name, slot_type, plugin, database FROM pg_catalog.pg_replication_slots')
+            for r in cursor:
+                value = {'type': r[1]}
+                if r[1] == 'logical':
+                    value.update({'plugin': r[2], 'database': r[3]})
+                replication_slots[r[0]] = value
+            self._replication_slots = replication_slots
             self._schedule_load_slots = False
 
     def postmaster_start_time(self):
         try:
-            cursor = self.query("""SELECT to_char(pg_postmaster_start_time(), 'YYYY-MM-DD HH24:MI:SS.MS TZ')""")
+            cursor = self.query("SELECT pg_catalog.to_char(pg_catalog.pg_postmaster_start_time(),"
+                                " 'YYYY-MM-DD HH24:MI:SS.MS TZ')")
             return cursor.fetchone()[0]
         except psycopg2.Error:
             return None
 
+    def drop_replication_slot(self, name):
+        cursor = self._query(('SELECT pg_catalog.pg_drop_replication_slot(%s) WHERE EXISTS (SELECT 1 ' +
+                              'FROM pg_catalog.pg_replication_slots WHERE slot_name = %s AND NOT active)'), name, name)
+        # In normal situation rowcount should be 1, otherwise either slot doesn't exists or it is still active
+        return cursor.rowcount == 1
+
+    @staticmethod
+    def compare_slots(s1, s2):
+        return s1['type'] == s2['type'] and\
+                (s1['type'] == 'physical' or s1['database'] == s2['database'] and s1['plugin'] == s2['plugin'])
+
     def sync_replication_slots(self, cluster):
         if self.use_slots:
             try:
                 self.load_replication_slots()
-                # if the replicatefrom tag is set on the member - we should not create the replication slot for it on
-                # the current master, because that member would replicate from elsewhere. We still create the slot if
-                # the replicatefrom destination member is currently not a member of the cluster (fallback to the
-                # master), or if replicatefrom destination member happens to be the current master
-                if self.role in ('master', 'standby_leader'):
-                    slot_members = [m.name for m in cluster.members if m.name != self.name and
-                                    (m.replicatefrom is None or m.replicatefrom == self.name or
-                                     not cluster.has_member(m.replicatefrom))]
-                else:
-                    # only manage slots for replicas that replicate from this one, except for the leader among them
-                    slot_members = [m.name for m in cluster.members if m.replicatefrom == self.name and
-                                    m.name != cluster.leader.name]
-                slots = set(slot_name_from_member_name(name) for name in slot_members)
 
-                if len(slots) < len(slot_members):
-                    # Find which names are conflicting for a nicer error message
-                    slot_conflicts = defaultdict(list)
-                    for name in slot_members:
-                        slot_conflicts[slot_name_from_member_name(name)].append(name)
-                    logger.error("Following cluster members share a replication slot name: %s",
-                                 "; ".join("{} map to {}".format(", ".join(v), k)
-                                           for k, v in slot_conflicts.items() if len(v) > 1))
+                slots = cluster.get_replication_slots(self.name, self.role)
 
-                # drop unused slots
-                for slot in set(self._replication_slots) - slots:
-                    cursor = self._query("""SELECT pg_drop_replication_slot(%s)
-                                             WHERE EXISTS(SELECT 1 FROM pg_replication_slots
-                                             WHERE slot_name = %s AND NOT active)""", slot, slot)
-
-                    if cursor.rowcount != 1:  # Either slot doesn't exists or it is still active
-                        self._schedule_load_slots = True  # schedule load_replication_slots on the next iteration
+                # drop old replication slots which are not presented in desired slots
+                for name in set(self._replication_slots) - set(slots):
+                    if not self.drop_replication_slot(name):
+                        logger.error("Failed to drop replication slot '%s'", name)
+                        self._schedule_load_slots = True
 
                 immediately_reserve = ', true' if self._major_version >= 90600 else ''
 
-                # create new slots
-                for slot in slots - set(self._replication_slots):
-                    self._query("""SELECT pg_create_physical_replication_slot(%s{0})
-                                    WHERE NOT EXISTS (SELECT 1 FROM pg_replication_slots
-                                    WHERE slot_name = %s)""".format(immediately_reserve), slot, slot)
+                logical_slots = defaultdict(dict)
+                for name, value in slots.items():
+                    if name in self._replication_slots and not self.compare_slots(value, self._replication_slots[name]):
+                        logger.info("Trying to drop replication slot '%s' because value is changing from %s to %s",
+                                    name, self._replication_slots[name], value)
+                        if not self.drop_replication_slot(name):
+                            logger.error("Failed to drop replication slot '%s'", name)
+                            self._schedule_load_slots = True
+                            continue
+                        self._replication_slots.pop(name)
+                    if name not in self._replication_slots:
+                        if value['type'] == 'physical':
+                            try:
+                                self._query(("SELECT pg_catalog.pg_create_physical_replication_slot(%s{0})" +
+                                             " WHERE NOT EXISTS (SELECT 1 FROM pg_catalog.pg_replication_slots" +
+                                             " WHERE slot_type = 'physical' AND slot_name = %s)").format(
+                                                 immediately_reserve), name, name)
+                            except Exception:
+                                logger.exception("Failed to create physical replication slot '%s'", name)
+                                self._schedule_load_slots = True
+                        elif value['type'] == 'logical' and name not in self._replication_slots:
+                            logical_slots[value['database']][name] = value
 
+                # create new logical slots
+                for database, values in logical_slots.items():
+                    conn_kwargs = self._local_connect_kwargs
+                    conn_kwargs['database'] = database
+                    with self._get_connection_cursor(**conn_kwargs) as cur:
+                        for name, value in values.items():
+                            try:
+                                cur.execute("SELECT pg_catalog.pg_create_logical_replication_slot(%s, %s)" +
+                                            " WHERE NOT EXISTS (SELECT 1 FROM pg_catalog.pg_replication_slots" +
+                                            " WHERE slot_type = 'logical' AND slot_name = %s)",
+                                            (name, value['plugin'], name))
+                            except Exception:
+                                logger.exception("Failed to create logical replication slot '%s' plugin='%s'",
+                                                 name, value['plugin'])
+                                self._schedule_load_slots = True
                 self._replication_slots = slots
             except Exception:
                 logger.exception('Exception when changing replication slots')
@@ -1753,9 +1774,9 @@ $$""".format(name, ' '.join(options)), name, password, password)
         # Pick candidates based on who has flushed WAL farthest.
         # TODO: for synchronous_commit = remote_write we actually want to order on write_location
         for app_name, state, sync_state in self.query(
-                """SELECT LOWER(application_name), state, sync_state
-                     FROM pg_stat_replication
-                    ORDER BY flush_{0} DESC""".format(self.lsn_name)):
+                "SELECT pg_catalog.lower(application_name), state, sync_state"
+                " FROM pg_catalog.pg_stat_replication"
+                " ORDER BY flush_{0} DESC".format(self.lsn_name)):
             member = members.get(app_name)
             if state != 'streaming' or not member or member.tags.get('nosync', False):
                 continue

patroni/scripts/wale_restore.py

@@ -224,13 +224,12 @@ class WALERestore(object):
                             lsn_name = 'location'
                         con.autocommit = True
                         with con.cursor() as cur:
-                            cur.execute("""SELECT CASE WHEN pg_is_in_recovery()
-                                                       THEN GREATEST(
-                                                                pg_{0}_{1}_diff(COALESCE(
-                                                                  pg_last_{0}_receive_{1}(), '0/0'), %s)::bigint,
-                                                                pg_{0}_{1}_diff(pg_last_{0}_replay_{1}(), %s)::bigint)
-                                                       ELSE pg_{0}_{1}_diff(pg_current_{0}_{1}(), %s)::bigint
-                                                   END""".format(wal_name, lsn_name),
+                            cur.execute(("SELECT CASE WHEN pg_catalog.pg_is_in_recovery()"
+                                         " THEN GREATEST(pg_catalog.pg_{0}_{1}_diff(COALESCE("
+                                         "pg_last_{0}_receive_{1}(), '0/0'), %s)::bigint, "
+                                         "pg_catalog.pg_{0}_{1}_diff(pg_catalog.pg_last_{0}_replay_{1}(), %s)::bigint)"
+                                         " ELSE pg_catalog.pg_{0}_{1}_diff(pg_catalog.pg_current_{0}_{1}(), %s)::bigint"
+                                         " END").format(wal_name, lsn_name),
                                         (backup_start_lsn, backup_start_lsn, backup_start_lsn))
 
                             diff_in_bytes = int(cur.fetchone()[0])

patroni/version.py

@@ -1 +1 @@
-__version__ = '1.5.0'
+__version__ = '1.5.2'

tests/test_api.py

@@ -95,6 +95,10 @@ class MockHa(object):
     def is_paused():
         return True
 
+    @staticmethod
+    def is_standby_cluster():
+        return False
+
 
 class MockPatroni(object):
 
@@ -167,8 +171,8 @@ class TestRestApiHandler(unittest.TestCase):
         self.assertIsNotNone(MockRestApiServer(RestApiHandler, 'GET /master'))
         with patch.object(RestApiServer, 'query', Mock(return_value=[('', 1, '', '', '', '', False, '')])):
             self.assertIsNotNone(MockRestApiServer(RestApiHandler, 'GET /patroni'))
-        MockPatroni.config.is_standby_cluster = PropertyMock(return_value=True)
-        MockRestApiServer(RestApiHandler, 'GET /standby_leader')
+        with patch.object(MockHa, 'is_standby_cluster', Mock(return_value=True)):
+            MockRestApiServer(RestApiHandler, 'GET /standby_leader')
 
     def test_do_OPTIONS(self):
         self.assertIsNotNone(MockRestApiServer(RestApiHandler, 'OPTIONS / HTTP/1.0'))

tests/test_ctl.py

@@ -192,24 +192,24 @@ class TestCtl(unittest.TestCase):
 
     def test_query_member(self):
         with patch('patroni.ctl.get_cursor', Mock(return_value=MockConnect().cursor())):
-            rows = query_member(None, None, None, 'master', 'SELECT pg_is_in_recovery()', {})
+            rows = query_member(None, None, None, 'master', 'SELECT pg_catalog.pg_is_in_recovery()', {})
             self.assertTrue('False' in str(rows))
 
-            rows = query_member(None, None, None, 'replica', 'SELECT pg_is_in_recovery()', {})
+            rows = query_member(None, None, None, 'replica', 'SELECT pg_catalog.pg_is_in_recovery()', {})
             self.assertEqual(rows, (None, None))
 
             with patch('test_postgresql.MockCursor.execute', Mock(side_effect=OperationalError('bla'))):
-                rows = query_member(None, None, None, 'replica', 'SELECT pg_is_in_recovery()', {})
+                rows = query_member(None, None, None, 'replica', 'SELECT pg_catalog.pg_is_in_recovery()', {})
 
         with patch('patroni.ctl.get_cursor', Mock(return_value=None)):
-            rows = query_member(None, None, None, None, 'SELECT pg_is_in_recovery()', {})
+            rows = query_member(None, None, None, None, 'SELECT pg_catalog.pg_is_in_recovery()', {})
             self.assertTrue('No connection to' in str(rows))
 
-            rows = query_member(None, None, None, 'replica', 'SELECT pg_is_in_recovery()', {})
+            rows = query_member(None, None, None, 'replica', 'SELECT pg_catalog.pg_is_in_recovery()', {})
             self.assertTrue('No connection to' in str(rows))
 
         with patch('patroni.ctl.get_cursor', Mock(side_effect=OperationalError('bla'))):
-            rows = query_member(None, None, None, 'replica', 'SELECT pg_is_in_recovery()', {})
+            rows = query_member(None, None, None, 'replica', 'SELECT pg_catalog.pg_is_in_recovery()', {})
 
     @patch('patroni.ctl.get_dcs')
     def test_dsn(self, mock_get_dcs):

tests/test_ha.py

@@ -5,7 +5,6 @@ import unittest
 import sys
 
 from mock import Mock, MagicMock, PropertyMock, patch
-from patroni.async_executor import CriticalTask
 from patroni.config import Config
 from patroni.dcs import Cluster, ClusterConfig, Failover, Leader, Member, get_dcs, SyncState, TimelineHistory
 from patroni.dcs.etcd import Client
@@ -62,17 +61,6 @@ def get_cluster_initialized_with_only_leader(failover=None, cluster_config=None)
     return get_cluster(True, leader, [leader], failover, None, cluster_config)
 
 
-def get_cluster_not_initialized_standby(failover=None, sync=None):
-    return get_cluster_not_initialized_without_leader(
-        cluster_config=ClusterConfig(1, {
-            "standby_cluster": {
-                "host": "localhost",
-                "port": 5432,
-                "primary_slot_name": "",
-            }}, 1)
-    )
-
-
 def get_standby_cluster_initialized_with_only_leader(failover=None, sync=None):
     return get_cluster_initialized_with_only_leader(
         cluster_config=ClusterConfig(1, {
@@ -158,7 +146,6 @@ def run_async(self, func, args=()):
 @patch.object(Postgresql, 'write_recovery_conf', Mock())
 @patch.object(Postgresql, 'query', Mock())
 @patch.object(Postgresql, 'checkpoint', Mock())
-@patch.object(Postgresql, 'call_nowait', Mock())
 @patch.object(Postgresql, 'cancellable_subprocess_call', Mock(return_value=0))
 @patch.object(etcd.Client, 'write', etcd_write)
 @patch.object(etcd.Client, 'read', etcd_read)
@@ -214,17 +201,10 @@ class TestHa(unittest.TestCase):
     @patch('patroni.dcs.etcd.Etcd.initialize', return_value=True)
     def test_start_as_standby_leader(self, initialize):
         self.p.data_directory_empty = true
-        self.ha.cluster = get_cluster_not_initialized_standby()
+        self.ha.cluster = get_cluster_not_initialized_without_leader(cluster_config=ClusterConfig(0, {}, 0))
         self.ha.cluster.is_unlocked = true
-        self.ha.patroni.config._dynamic_configuration = {"standby_cluster": {
-            "host": "localhost",
-            "port": 5432,
-            "primary_slot_name": "",
-        }}
-        self.assertEqual(
-            self.ha.run_cycle(),
-            'trying to bootstrap a new standby leader'
-        )
+        self.ha.patroni.config._dynamic_configuration = {"standby_cluster": {"port": 5432}}
+        self.assertEqual(self.ha.run_cycle(), 'trying to bootstrap a new standby leader')
 
     @patch.object(Cluster, 'get_clone_member',
                   Mock(return_value=Member(0, 'test', 1, {'api_url': 'http://127.0.0.1:8011/patroni'})))
@@ -233,27 +213,7 @@ class TestHa(unittest.TestCase):
         self.p.data_directory_empty = true
         self.ha.cluster = get_standby_cluster_initialized_with_only_leader()
         self.ha.cluster.is_unlocked = false
-        self.ha.patroni.config._dynamic_configuration = {"standby_cluster": {
-            "host": "localhost",
-            "port": 5432,
-            "primary_slot_name": "",
-        }}
-        self.assertEqual(
-            self.ha.run_cycle(),
-            "trying to bootstrap from replica 'test'"
-        )
-
-    @patch.object(Postgresql, 'create_replica', Mock(return_value=0))
-    def test_bootstrap_standby_leader(self):
-        self.ha.cluster = get_cluster_not_initialized_standby()
-        self.ha.cluster.is_unlocked = true
-        self.ha.patroni.config._dynamic_configuration = {"standby_cluster": {
-            "host": "localhost",
-            "port": 5432,
-            "primary_slot_name": "",
-        }}
-        self.ha._post_bootstrap_task = CriticalTask()
-        self.assertEqual(self.ha.bootstrap_standby_leader(), True)
+        self.assertEqual(self.ha.run_cycle(), "trying to bootstrap from replica 'test'")
 
     def test_recover_replica_failed(self):
         self.p.controldata = lambda: {'Database cluster state': 'in recovery', 'Database system identifier': SYSID}
@@ -692,11 +652,6 @@ class TestHa(unittest.TestCase):
     def test_process_healthy_standby_cluster_as_standby_leader(self):
         self.p.is_leader = false
         self.p.name = 'leader'
-        self.ha.patroni.config._dynamic_configuration = {"standby_cluster": {
-            "host": "localhost",
-            "port": 5432,
-            "primary_slot_name": "",
-        }}
         self.ha.cluster = get_standby_cluster_initialized_with_only_leader()
         msg = 'no action.  i am the standby leader with the lock'
         self.assertEqual(self.ha.run_cycle(), msg)
@@ -704,24 +659,13 @@ class TestHa(unittest.TestCase):
     def test_process_healthy_standby_cluster_as_cascade_replica(self):
         self.p.is_leader = false
         self.p.name = 'replica'
-        self.ha.patroni.config._dynamic_configuration = {"standby_cluster": {
-            "host": "localhost",
-            "port": 5432,
-            "primary_slot_name": "",
-        }}
         self.ha.cluster = get_standby_cluster_initialized_with_only_leader()
         msg = 'no action.  i am a secondary and i am following a leader'
         self.assertEqual(self.ha.run_cycle(), msg)
 
-    @patch('patroni.dcs.etcd.Etcd.initialize', return_value=True)
-    def test_process_unhealthy_standby_cluster_as_standby_leader(self, initialize):
+    def test_process_unhealthy_standby_cluster_as_standby_leader(self):
         self.p.is_leader = false
         self.p.name = 'leader'
-        self.ha.patroni.config._dynamic_configuration = {"standby_cluster": {
-            "host": "localhost",
-            "port": 5432,
-            "primary_slot_name": "",
-        }}
         self.ha.cluster = get_standby_cluster_initialized_with_only_leader()
         self.ha.cluster.is_unlocked = true
         self.ha.sysid_valid = true
@@ -730,19 +674,30 @@ class TestHa(unittest.TestCase):
         self.assertEqual(self.ha.run_cycle(), msg)
 
     @patch.object(Postgresql, 'rewind_needed_and_possible', Mock(return_value=True))
-    @patch('patroni.dcs.etcd.Etcd.initialize', return_value=True)
-    def test_process_unhealthy_standby_cluster_as_cascade_replica(self, initialize):
+    def test_process_unhealthy_standby_cluster_as_cascade_replica(self):
         self.p.is_leader = false
         self.p.name = 'replica'
-        self.ha.patroni.config._dynamic_configuration = {"standby_cluster": {
-            "host": "localhost",
-            "port": 5432,
-            "primary_slot_name": "",
-        }}
         self.ha.cluster = get_standby_cluster_initialized_with_only_leader()
         self.ha.is_unlocked = true
-        msg = 'running pg_rewind from leader'
-        self.assertEqual(self.ha.run_cycle(), msg)
+        self.assertTrue(self.ha.run_cycle().startswith('running pg_rewind from remote_master:'))
+
+    def test_recover_unhealthy_leader_in_standby_cluster(self):
+        self.p.is_leader = false
+        self.p.name = 'leader'
+        self.p.is_running = false
+        self.p.follow = false
+        self.ha.cluster = get_standby_cluster_initialized_with_only_leader()
+        self.assertEqual(self.ha.run_cycle(), 'starting as a standby leader because i had the session lock')
+
+    def test_recover_unhealthy_unlocked_standby_cluster(self):
+        self.p.is_leader = false
+        self.p.name = 'leader'
+        self.p.is_running = false
+        self.p.follow = false
+        self.ha.cluster = get_standby_cluster_initialized_with_only_leader()
+        self.ha.cluster.is_unlocked = true
+        self.ha.has_lock = false
+        self.assertEqual(self.ha.run_cycle(), 'trying to follow a remote master because standby cluster is unhealthy')
 
     def test_failed_to_update_lock_in_pause(self):
         self.ha.update_lock = false
@@ -1045,3 +1000,12 @@ class TestHa(unittest.TestCase):
         self.assertEqual(self.ha.run_cycle(), 'PAUSE: no action.  i am the leader with the lock')
         self.ha.is_paused = false
         self.assertEqual(self.ha.run_cycle(), 'no action.  i am the leader with the lock')
+
+    @patch('psycopg2.connect', psycopg2_connect)
+    def test_permanent_logical_slots_after_promote(self):
+        config = ClusterConfig(1, {'slots': {'l': {'database': 'postgres', 'plugin': 'test_decoding'}}}, 1)
+        self.ha.cluster = get_cluster_initialized_without_leader(cluster_config=config)
+        self.assertEqual(self.ha.run_cycle(), 'acquired session lock as a leader')
+        self.ha.cluster = get_cluster_initialized_without_leader(leader=True, cluster_config=config)
+        self.ha.has_lock = true
+        self.assertEqual(self.ha.run_cycle(), 'no action.  i am the leader with the lock')

tests/test_kubernetes.py

@@ -50,6 +50,13 @@ class TestKubernetes(unittest.TestCase):
                         'labels': {'f': 'b'}, 'use_endpoints': True, 'pod_ip': '10.0.0.0'})
         self.assertIsNotNone(k.update_leader('123'))
 
+    @patch('kubernetes.config.load_kube_config', Mock())
+    @patch.object(k8s_client.CoreV1Api, 'create_namespaced_endpoints', Mock())
+    def test_update_leader_with_restricted_access(self):
+        k = Kubernetes({'ttl': 30, 'scope': 'test', 'name': 'p-0', 'retry_timeout': 10,
+                        'labels': {'f': 'b'}, 'use_endpoints': True, 'pod_ip': '10.0.0.0'})
+        self.assertIsNotNone(k.update_leader('123', True))
+
     def test_take_leader(self):
         self.k.take_leader()
         self.k._leader_observed_record['leader'] = 'test'

tests/test_postgresql.py

@@ -8,7 +8,7 @@ import unittest
 
 from mock import Mock, MagicMock, PropertyMock, patch, mock_open
 from patroni.async_executor import CriticalTask
-from patroni.dcs import Cluster, Leader, Member, RemoteMember, SyncState
+from patroni.dcs import Cluster, ClusterConfig, Leader, Member, RemoteMember, SyncState
 from patroni.exceptions import PostgresConnectionException, PostgresException
 from patroni.postgresql import Postgresql, STATE_REJECT, STATE_NO_RESPONSE
 from patroni.postmaster import PostmasterProcess
@@ -28,15 +28,15 @@ class MockCursor(object):
     def execute(self, sql, *params):
         if sql.startswith('blabla'):
             raise psycopg2.ProgrammingError()
-        elif sql == 'CHECKPOINT':
+        elif sql == 'CHECKPOINT' or sql.startswith('SELECT pg_catalog.pg_create_'):
             raise psycopg2.OperationalError()
         elif sql.startswith('RetryFailedError'):
             raise RetryFailedError('retry')
         elif sql.startswith('SELECT slot_name'):
-            self.results = [('blabla',), ('foobar',)]
-        elif sql.startswith('SELECT CASE WHEN pg_is_in_recovery()'):
+            self.results = [('blabla', 'physical'), ('foobar', 'physical'), ('ls', 'logical', 'a', 'b')]
+        elif sql.startswith('SELECT CASE WHEN pg_catalog.pg_is_in_recovery()'):
             self.results = [(1, 2)]
-        elif sql.startswith('SELECT pg_is_in_recovery()'):
+        elif sql.startswith('SELECT pg_catalog.pg_is_in_recovery()'):
             self.results = [(False, 2)]
         elif sql.startswith('WITH replication_info AS ('):
             replication_info = '[{"application_name":"walreceiver","client_addr":"1.2.3.4",' +\
@@ -53,7 +53,7 @@ class MockCursor(object):
             self.results = [('1', 2, '0/402EEC0', '')]
         elif sql.startswith('SELECT isdir, modification'):
             self.results = [(False, datetime.datetime.now())]
-        elif sql.startswith('SELECT pg_read_file'):
+        elif sql.startswith('SELECT pg_catalog.pg_read_file'):
             self.results = [('1\t0/40159C0\tno recovery target specified\n\n' +
                              '2\t1/40159C0\tno recovery target specified\n',)]
         elif sql.startswith('TIMELINE_HISTORY '):
@@ -421,9 +421,13 @@ class TestPostgresql(unittest.TestCase):
     def test_create_replica(self, mock_cancellable_subprocess_call):
         self.p.delete_trigger_file = Mock(side_effect=OSError)
 
+        self.p.config['create_replica_methods'] = ['pgBackRest']
+        self.p.config['pgBackRest'] = {'command': 'pgBackRest', 'keep_data': True, 'no_params': True}
+        mock_cancellable_subprocess_call.return_value = 0
+        self.assertEqual(self.p.create_replica(self.leader), 0)
+
         self.p.config['create_replica_methods'] = ['wale', 'basebackup']
         self.p.config['wale'] = {'command': 'foo'}
-        mock_cancellable_subprocess_call.return_value = 0
         self.assertEqual(self.p.create_replica(self.leader), 0)
         del self.p.config['wale']
         self.assertEqual(self.p.create_replica(self.leader), 0)
@@ -498,23 +502,25 @@ class TestPostgresql(unittest.TestCase):
     @patch.object(Postgresql, 'is_running', Mock(return_value=True))
     def test_sync_replication_slots(self):
         self.p.start()
-        cluster = Cluster(True, None, self.leader, 0, [self.me, self.other, self.leadermem], None, None, None)
+        config = ClusterConfig(1, {'slots': {'ls': {'database': 'a', 'plugin': 'b'},
+                                             'A': 0, 'test_3': 0, 'b': {'type': 'logical', 'plugin': '1'}}}, 1)
+        cluster = Cluster(True, config, self.leader, 0, [self.me, self.other, self.leadermem], None, None, None)
         with mock.patch('patroni.postgresql.Postgresql._query', Mock(side_effect=psycopg2.OperationalError)):
             self.p.sync_replication_slots(cluster)
         self.p.sync_replication_slots(cluster)
         with mock.patch('patroni.postgresql.Postgresql.role', new_callable=PropertyMock(return_value='replica')):
             self.p.sync_replication_slots(cluster)
-        with mock.patch('patroni.postgresql.logger.error', new_callable=Mock()) as errorlog_mock:
+        with patch.object(Postgresql, 'drop_replication_slot', Mock(return_value=True)),\
+                patch('patroni.dcs.logger.error', new_callable=Mock()) as errorlog_mock:
             self.p.query = Mock()
             alias1 = Member(0, 'test-3', 28, {'conn_url': 'postgres://replicator:[email protected]:5436/postgres'})
             alias2 = Member(0, 'test.3', 28, {'conn_url': 'postgres://replicator:[email protected]:5436/postgres'})
             cluster.members.extend([alias1, alias2])
             self.p.sync_replication_slots(cluster)
-            errorlog_mock.assert_called_once()
-            self.assertTrue("test-3" in errorlog_mock.call_args[0][1],
-                            "non matching {0}".format(errorlog_mock.call_args[0][1]))
-            self.assertTrue("test.3" in errorlog_mock.call_args[0][1],
-                            "non matching {0}".format(errorlog_mock.call_args[0][1]))
+            self.assertEqual(errorlog_mock.call_count, 5)
+            ca = errorlog_mock.call_args_list[0][0][1]
+            self.assertTrue("test-3" in ca, "non matching {0}".format(ca))
+            self.assertTrue("test.3" in ca, "non matching {0}".format(ca))
 
     @patch.object(MockCursor, 'execute', Mock(side_effect=psycopg2.OperationalError))
     def test__query(self):

tests/test_zookeeper.py

@@ -113,6 +113,7 @@ class TestPatroniSequentialThreadingHandler(unittest.TestCase):
     def test_create_connection(self):
         self.assertIsNotNone(self.handler.create_connection(()))
         self.assertIsNotNone(self.handler.create_connection((), 40))
+        self.assertIsNotNone(self.handler.create_connection(timeout=40))
 
 
 class TestZooKeeper(unittest.TestCase):