Ian Barwick
b900f9996f
"repmgr standby clone": add --dry-run option
2017-08-28 15:04:50 +09:00
Ian Barwick
e05bab8284
"standby switchover": epxlictly confirm suitability for --pg-rewind
...
If --force-rewind requested.
2017-08-28 14:50:08 +09:00
Ian Barwick
4ba2291cdd
Update README
2017-08-28 12:32:05 +09:00
Ian Barwick
3b28cfa70e
Update "standby switchover" command reference
2017-08-28 11:58:10 +09:00
Ian Barwick
4b2fb7d15b
Update "repmgr primary ..." --help output and command reference
2017-08-28 11:53:41 +09:00
Ian Barwick
d85b066f45
repmgrd cluster show: add warnings with information about discrepancies
2017-08-28 11:19:32 +09:00
Ian Barwick
4a11551c2f
repmgrd: handle local node failure
2017-08-28 10:31:43 +09:00
Ian Barwick
754084c814
Update "repmgr standby --help" output
2017-08-26 10:27:22 +09:00
Ian Barwick
57215a8bd7
Add --help output for "standby clone"
2017-08-25 23:07:17 +09:00
Ian Barwick
c1ed248fb1
Handle "event_notifications" when reloading configuration
2017-08-25 23:07:07 +09:00
Ian Barwick
e280e2670e
Add code to reload configuration file
2017-08-25 22:31:42 +09:00
Ian Barwick
5208655a35
Parse "recovery_min_apply_delay" from recovery.conf
2017-08-25 21:47:14 +09:00
Ian Barwick
5ee1eb6bf7
Convert --recovery-min-apply-delay to configuration file option
...
That way it only needs to be set once, and won't get lost during
follow operations etc.
2017-08-25 21:25:15 +09:00
Ian Barwick
dc172cae20
When performing a follow operation, start/restart server as appropriate
...
Before this we were always forcing a restart, which is technically not
a problem but produces some potentially confusing log entries along the
lines:
pg_ctl: PID file "/path/to/postmaster.pid" does not exist
Is server running?
starting server anyway
2017-08-25 16:50:30 +09:00
Ian Barwick
1ba2a2462c
Update README
2017-08-25 16:48:14 +09:00
Ian Barwick
4921f389b6
Fix spurious warning when executing "repmgr node rejoin"
...
Database connection parameters required for this.
2017-08-25 16:28:25 +09:00
Ian Barwick
a449e8512e
repmgr: improve "repmgr standby switchover" log output
...
Particularly in --dry-run mode it's useful to get a confirmation that
various prerequisites are met.
2017-08-25 16:01:11 +09:00
Ian Barwick
1452238613
Update document "repmgrd-node-fencing.md"
2017-08-25 15:58:52 +09:00
Ian Barwick
3b9d201e18
Add/update document "repmgrd-node-fencing.md"
2017-08-25 12:32:55 +09:00
Ian Barwick
2092a55b9e
Update README
...
Document "standby switchover" and additional repmgrd information.
2017-08-25 00:39:22 +09:00
Ian Barwick
fcd111ac4c
Improve logging output during failover process
2017-08-24 22:44:03 +09:00
Ian Barwick
7a00ad930a
Update README
2017-08-24 22:43:55 +09:00
Ian Barwick
db157ad9bc
Update README
2017-08-24 17:43:01 +09:00
Ian Barwick
7d2dc0aa89
Update README
2017-08-24 16:51:22 +09:00
Ian Barwick
8a8982d7d9
Update README
2017-08-24 16:09:13 +09:00
Ian Barwick
4f86fd681c
Update repmgr--unpackaged--4.0.sql
2017-08-24 15:41:54 +09:00
Ian Barwick
eee8d65259
Update view "replication_status"
2017-08-24 15:05:13 +09:00
Ian Barwick
a127e8face
Provisionally add view "repmgr.replication_status"
2017-08-24 12:58:36 +09:00
Ian Barwick
a659132ea4
repmgrd: write monitoring statistics
2017-08-24 11:49:44 +09:00
Ian Barwick
5dfb8a5b06
Add SQL for monitoring history table
2017-08-23 22:18:23 +09:00
Ian Barwick
8dfb7bbc7d
repmgrd: handle promotion failure properly
2017-08-23 21:44:18 +09:00
Ian Barwick
6259463007
repmgrd: various fixes for "manual" failover mode
2017-08-23 10:56:55 +09:00
Ian Barwick
ff07763242
repmgr: update --help output
...
Display database connection options.
2017-08-22 15:07:22 +09:00
Ian Barwick
ef0163bd84
"standby follow": ensure recovery.conf uses "node_name" as "application_name"
...
In repmgr4 we want to make it easier to establish which node is connected
to which.
2017-08-22 13:21:29 +09:00
Ian Barwick
44d1b96bc3
repmgrd: fix log output
2017-08-22 13:21:16 +09:00
Ian Barwick
7ca396b9cb
Add missing Barman options check
2017-08-21 14:10:08 +09:00
Ian Barwick
4943909282
Fix source server version number checks during "standby clone"
2017-08-21 13:36:11 +09:00
Ian Barwick
f131d37871
Update document "changes-in-repmgr4.md"
2017-08-21 12:53:47 +09:00
Ian Barwick
5df77f0fed
Fixes to contrib/convert-config.pl
2017-08-21 10:15:04 +09:00
Ian Barwick
594e9e5007
Document upgrade process from repmgr3
...
Also provide unpackaged extension upgrade SQL, and a script to
assist converting repmgr.conf files.
2017-08-17 23:37:31 +09:00
Ian Barwick
5e019a65e7
Initial upgrade SQL for repmgr3 to repmgr4
2017-08-17 21:14:55 +09:00
Ian Barwick
f3579e5fb0
Update README
2017-08-17 16:06:57 +09:00
Ian Barwick
0572996d61
Update README
2017-08-17 15:36:21 +09:00
Ian Barwick
da24d883e5
Remove option "--wal-keep-segments"
...
This is a remnant of the early repmgr days when there were no alternative
mechanisms for ensuring sufficient WAL remains available while cloning a
standby.
The purpose of this setting was to override a check for an (arbitrary)
minimum setting for "wal_keep_segments". As there's no reliable way
of determining a sensible value for this, and improvements in
pg_basebackup mean WALs can be streamed (possibly using a replication
slot) while the backup is in progress, there's no point in keeping
this around.
We will however still emit a warning about setting "wal_keep_segments"
if the configuration doesn't appear to provide any other way of
ensuring WAL is available during/after the cloning process and
"wal_keep_segments" is not set.
2017-08-17 14:45:13 +09:00
Ian Barwick
ac64526bd3
Finalize "repmgr node --help" output
2017-08-17 14:25:43 +09:00
Ian Barwick
9ac5c48cc5
Add --help output for "repmgr node check"
2017-08-17 14:12:40 +09:00
Ian Barwick
5127b9253b
"repmgr node check": improve Nagios output
2017-08-17 13:54:48 +09:00
Ian Barwick
b1ba476241
Rename "archiver" check etc. to "archive-ready"
...
Gives a better indication of what's being checked.
2017-08-17 12:23:56 +09:00
Ian Barwick
b1b5870d54
"repmgr node status": add --help output, fix CSV output
...
Also ensure is executed only on local node, as it needs to read
the data directory.
2017-08-17 11:27:31 +09:00
Ian Barwick
7a9064cd1b
"repmgr cluster events": show node name in output, if available
...
Nodes can be removed from repmgr.nodes, so we'll only have the historical
ID available via repmgr.events.
2017-08-17 10:49:54 +09:00
