mirror of
https://github.com/EnterpriseDB/repmgr.git
synced 2026-03-23 15:16:29 +00:00
302 lines
11 KiB
Markdown
302 lines
11 KiB
Markdown
repmgr: Replication Manager for PostgreSQL
|
|
==========================================
|
|
|
|
`repmgr` is a suite of open-source tools to manage replication and failover
|
|
within a cluster of PostgreSQL servers. It enhances PostgreSQL's built-in
|
|
replication capabilities with utilities to set up standby servers, monitor
|
|
replication, and perform administrative tasks such as failover or switchover
|
|
operations.
|
|
|
|
`repmgr 4` is a complete rewrite of the existing `repmgr` codebase.
|
|
|
|
Supports PostgreSQL 9.5 and later; support for PostgreSQL 9.3 and 9.4 has been
|
|
dropped. To use `repmgr 4` with BDR 2.0, PostgreSQL 9.6 is required.
|
|
|
|
Building from source
|
|
--------------------
|
|
|
|
Simply:
|
|
|
|
./configure && make install
|
|
|
|
Ensure `pg_config` for the target PostgreSQL version is in `$PATH`.
|
|
|
|
|
|
Reference
|
|
---------
|
|
|
|
### repmgr commands
|
|
|
|
The following commands are available:
|
|
|
|
repmgr primary register
|
|
repmgr primary unregister
|
|
|
|
repmgr standby clone
|
|
repmgr standby register
|
|
repmgr standby unregister
|
|
repmgr standby promote
|
|
repmgr standby follow
|
|
|
|
repmgr bdr register
|
|
repmgr bdr unregister
|
|
|
|
repmgr node status
|
|
|
|
repmgr cluster show
|
|
repmgr cluster event [--all] [--node-id] [--node-name] [--event] [--event-matching]
|
|
|
|
|
|
* `primary register`
|
|
|
|
Registers a primary in a streaming replication cluster, and configures
|
|
it for use with repmgr. This command needs to be executed before any
|
|
standby nodes are registered.
|
|
|
|
`master register` can be used as an alias for `primary register`.
|
|
|
|
* `cluster show`
|
|
|
|
Displays information about each active node in the replication cluster. This
|
|
command polls each registered server and shows its role (`primary` / `standby` /
|
|
`bdr`) and status. It polls each server directly and can be run on any node
|
|
in the cluster; this is also useful when analyzing connectivity from a particular
|
|
node.
|
|
|
|
This command requires either a valid `repmgr.conf` file or a database connection
|
|
string to one of the registered nodes; no additional arguments are needed.
|
|
|
|
Example:
|
|
|
|
$ repmgr -f /etc/repmgr.conf cluster show
|
|
|
|
ID | Name | Role | Status | Upstream | Connection string
|
|
----+-------+---------+-----------+----------+-----------------------------------------
|
|
1 | node1 | primary | * running | | host=db_node1 dbname=repmgr user=repmgr
|
|
2 | node2 | standby | running | node1 | host=db_node2 dbname=repmgr user=repmgr
|
|
3 | node3 | standby | running | node1 | host=db_node3 dbname=repmgr user=repmgr
|
|
|
|
To show database connection errors when polling nodes, run the command in
|
|
`--verbose` mode.
|
|
|
|
The `cluster show` command accepts an optional parameter `--csv`, which
|
|
outputs the replication cluster's status in a simple CSV format, suitable for
|
|
parsing by scripts:
|
|
|
|
$ repmgr -f /etc/repmgr.conf cluster show --csv
|
|
1,-1,-1
|
|
2,0,0
|
|
3,0,1
|
|
|
|
The columns have following meanings:
|
|
|
|
- node ID
|
|
- availability (0 = available, -1 = unavailable)
|
|
- recovery state (0 = not in recovery, 1 = in recovery, -1 = unknown)
|
|
|
|
Note that the availability is tested by connecting from the node where
|
|
`repmgr cluster show` is executed, and does not necessarily imply the node
|
|
is down.
|
|
|
|
|
|
* `cluster matrix` and `cluster crosscheck`
|
|
|
|
These commands display connection information for each pair of
|
|
nodes in the replication cluster.
|
|
|
|
- `cluster matrix` runs a `cluster show` on each node and arranges
|
|
the results in a matrix, recording success or failure;
|
|
|
|
- `cluster crosscheck` runs a `cluster matrix` on each node and
|
|
combines the results in a single matrix, providing a full
|
|
overview of connections between all databases in the cluster.
|
|
|
|
These commands require a valid `repmgr.conf` file on each node.
|
|
Additionally passwordless `ssh` connections are required between
|
|
all nodes.
|
|
|
|
Example 1 (all nodes up):
|
|
|
|
$ repmgr -f /etc/repmgr.conf cluster matrix
|
|
|
|
Name | Id | 1 | 2 | 3
|
|
-------+----+----+----+----
|
|
node1 | 1 | * | * | *
|
|
node2 | 2 | * | * | *
|
|
node3 | 3 | * | * | *
|
|
|
|
Here `cluster matrix` is sufficient to establish the state of each
|
|
possible connection.
|
|
|
|
|
|
Example 2 (node1 and `node2` up, `node3` down):
|
|
|
|
$ repmgr -f /etc/repmgr.conf cluster matrix
|
|
|
|
Name | Id | 1 | 2 | 3
|
|
-------+----+----+----+----
|
|
node1 | 1 | * | * | x
|
|
node2 | 2 | * | * | x
|
|
node3 | 3 | ? | ? | ?
|
|
|
|
Each row corresponds to one server, and indicates the result of
|
|
testing an outbound connection from that server.
|
|
|
|
Since `node3` is down, all the entries in its row are filled with
|
|
"?", meaning that there we cannot test outbound connections.
|
|
|
|
The other two nodes are up; the corresponding rows have "x" in the
|
|
column corresponding to node3, meaning that inbound connections to
|
|
that node have failed, and "*" in the columns corresponding to
|
|
node1 and node2, meaning that inbound connections to these nodes
|
|
have succeeded.
|
|
|
|
In this case, `cluster crosscheck` gives the same result as `cluster
|
|
matrix`, because from any functioning node we can observe the same
|
|
state: `node1` and `node2` are up, `node3` is down.
|
|
|
|
Example 3 (all nodes up, firewall dropping packets originating
|
|
from `node1` and directed to port 5432 on node3)
|
|
|
|
Running `cluster matrix` from `node1` gives the following output:
|
|
|
|
$ repmgr -f /etc/repmgr.conf cluster matrix
|
|
|
|
Name | Id | 1 | 2 | 3
|
|
-------+----+----+----+----
|
|
node1 | 1 | * | * | x
|
|
node2 | 2 | * | * | *
|
|
node3 | 3 | ? | ? | ?
|
|
|
|
(Note this may take some time depending on the `connect_timeout`
|
|
setting in the registered node `conninfo` strings; default is 1
|
|
minute which means without modification the above command would
|
|
take around 2 minutes to run; see comment elsewhere about setting
|
|
`connect_timeout`)
|
|
|
|
The matrix tells us that we cannot connect from `node1` to `node3`,
|
|
and that (therefore) we don't know the state of any outbound
|
|
connection from node3.
|
|
|
|
In this case, the `cluster crosscheck` command is more informative:
|
|
|
|
$ repmgr -f /etc/repmgr.conf cluster crosscheck
|
|
|
|
Name | Id | 1 | 2 | 3
|
|
-------+----+----+----+----
|
|
node1 | 1 | * | * | x
|
|
node2 | 2 | * | * | *
|
|
node3 | 3 | * | * | *
|
|
|
|
What happened is that `cluster crosscheck` merged its own `cluster
|
|
matrix` with the `cluster matrix` output from `node2`; the latter is
|
|
able to connect to `node3` and therefore determine the state of
|
|
outbound connections from that node.
|
|
|
|
|
|
Backwards compatibility
|
|
-----------------------
|
|
|
|
`repmgr` is now implemented as a PostgreSQL extension, and all database
|
|
objects used by repmgr are stored in a dedicated `repmgr` schema, rather
|
|
than `repmgr_$cluster_name`. Note there is no need to install the extension,
|
|
this will be done automatically by `repmgr primary register`.
|
|
|
|
Metadata tables have been revised and are not backwards-compatible
|
|
with repmgr 3.x. (however future DDL updates will be easier as they can be
|
|
carried out via the ALTER EXTENSION mechanism.).
|
|
|
|
An extension upgrade script will be provided for pre-4.0 installations;
|
|
note this will require the existing `repmgr_$cluster_name` schema to
|
|
be renamed to `repmgr` beforehand.
|
|
|
|
Some configuration items have had their names changed for consistency
|
|
and clarity e.g. `node` => `node_id`. `repmgr` will issue a warning
|
|
about deprecated/altered options.
|
|
|
|
Some configuration items have been changed to command line options,
|
|
and vice-versa, e.g. to avoid hard-coding items such as a a node's
|
|
upstream ID, which might change over time.
|
|
|
|
See file `doc/changes-in-repmgr4.md` for more details.
|
|
|
|
|
|
Generating event notifications with repmgr/repmgrd
|
|
--------------------------------------------------
|
|
|
|
Each time `repmgr` or `repmgrd` perform a significant event, a record
|
|
of that event is written into the `repmgr.events` table together with
|
|
a timestamp, an indication of failure or success, and further details
|
|
if appropriate. This is useful for gaining an overview of events
|
|
affecting the replication cluster. However note that this table has
|
|
advisory character and should be used in combination with the `repmgr`
|
|
and PostgreSQL logs to obtain details of any events.
|
|
|
|
Example output after a primary was registered and a standby cloned
|
|
and registered:
|
|
|
|
repmgr=# SELECT * from repmgr.events ;
|
|
node_id | event | successful | event_timestamp | details
|
|
---------+------------------+------------+-------------------------------+-------------------------------------------------------------------------------------
|
|
1 | primary_register | t | 2016-01-08 15:04:39.781733+09 |
|
|
2 | standby_clone | t | 2016-01-08 15:04:49.530001+09 | Cloned from host 'repmgr_node1', port 5432; backup method: pg_basebackup; --force: N
|
|
2 | standby_register | t | 2016-01-08 15:04:50.621292+09 |
|
|
(3 rows)
|
|
|
|
Alternatively use `repmgr cluster event` to output a list of events.
|
|
|
|
Additionally, event notifications can be passed to a user-defined program
|
|
or script which can take further action, e.g. send email notifications.
|
|
This is done by setting the `event_notification_command` parameter in
|
|
`repmgr.conf`.
|
|
|
|
This parameter accepts the following format placeholders:
|
|
|
|
%n - node ID
|
|
%e - event type
|
|
%s - success (1 or 0)
|
|
%t - timestamp
|
|
%d - details
|
|
|
|
The values provided for "%t" and "%d" will probably contain spaces,
|
|
so should be quoted in the provided command configuration, e.g.:
|
|
|
|
event_notification_command='/path/to/some/script %n %e %s "%t" "%d"'
|
|
|
|
Additionally the following format placeholders are available for the event
|
|
type `bdr_failover` and optionally `bdr_recovery`:
|
|
|
|
%c - conninfo string of the next available node
|
|
%a - name of the next available node
|
|
|
|
These should always be quoted.
|
|
|
|
By default, all notification type will be passed to the designated script;
|
|
the notification types can be filtered to explicitly named ones:
|
|
|
|
event_notifications=primary_register,standby_register
|
|
|
|
The following event types are available:
|
|
|
|
* `master_register`
|
|
* `standby_register`
|
|
* `standby_unregister`
|
|
* `standby_clone`
|
|
* `standby_promote`
|
|
* `standby_follow`
|
|
* `standby_disconnect_manual`
|
|
* `repmgrd_start`
|
|
* `repmgrd_shutdown`
|
|
* `repmgrd_failover_promote`
|
|
* `repmgrd_failover_follow`
|
|
* `bdr_failover`
|
|
* `bdr_reconnect`
|
|
* `bdr_recovery`
|
|
* `bdr_register`
|
|
* `bdr_unregister`
|
|
|
|
Note that under some circumstances (e.g. no replication cluster master could
|
|
be located), it will not be possible to write an entry into the `repmgr.events`
|
|
table, in which case executing a script via `event_notification_command` can
|
|
serve as a fallback by generating some form of notification.
|