github/repmgr
1
0
Fork 0
mirror of https://github.com/EnterpriseDB/repmgr.git synced 2026-03-23 15:16:29 +00:00
Code Issues Packages Projects Releases 1 Wiki Activity
1,328 Commits 43 Branches 68 Tags
9788b2bd29f154ef0a3775047889989ce5a650de
Commit Graph

1328 Commits

This Branch
This Branch
All Branches
Author SHA1 Message Date
Ian Barwick
9788b2bd29 repmgr: don't display timestamp in log output 
Differentiate between repmgr and repmgrd output
 2016-12-05 10:44:30 +09:00
Ian Barwick
227f0190f7 repmgr: initial support for PostgreSQL 10 
Handle directory name change from pg_xlog to pg_wal.

Note that some functions with "xlog" in their name may also change.
 2016-12-02 12:06:08 +09:00
Ian Barwick
d6dbc70916 Update HISTORY 2016-11-25 08:17:01 +09:00
Ian Barwick
d2f4eda224 Note use of phbouncer %include directive 2016-11-23 09:33:30 +09:00
Ian Barwick
2588853e83 Refactor sample script in repmgrd-node-fencing.md 2016-11-23 09:21:14 +09:00
Ian Barwick
b54f98ed8a repmgr: always log to STDERR even if log facility defined 2016-11-23 08:55:55 +09:00
Ian Barwick
26f73686e5 repmgrd: fixes to configuration reload mechanism 2016-11-02 23:23:26 +09:00
Ian Barwick
e274a2cbcb repmgrd: clean up some redundant code 2016-11-02 16:52:44 +09:00
Ian Barwick
d502bbe614 repmgrd: enable logging configuration to be changed 2016-11-01 20:34:50 +09:00
Ian Barwick
2594411820 reload_config(): document items which can change 2016-11-01 20:08:57 +09:00
Ian Barwick
d22535de00 repmgr: pointless parsing the configuration here 2016-11-01 19:28:03 +09:00
Ian Barwick
fce1f0cd4a Refactor reload_config() 
Remove any non-repmgrd specific items.

parse_config() already sanity-checks the values so no need to
recheck. Refactor parse_config() so when called by reload_config()
it won't exit if errors are encountered.
 2016-11-01 19:05:21 +09:00
Ian Barwick
bb842c3989 repmgr: improve replication status checking during switchover 
When checking the new standby's record in pg_stat_replication, keep
polling until the expected status is reported, and only give up
after a timeout was exceeded.

Previously repmgr would report an error if status was "startup",
even though this is not a problem.
 2016-11-01 17:42:35 +09:00
Ian Barwick
556ff3c311 repmgrd: clarify master_response_timeout 2016-11-01 15:56:57 +09:00
Ian Barwick
251486546d repmgr: when unregistering a witness, remove record on that witness to 
Fixes GitHub #250.
 2016-10-30 16:28:12 +09:00
Ian Barwick
53d3e71cd3 README: formatting 2016-10-27 10:42:42 +09:00
Ian Barwick
b986ce81b2 README: link to latest release notes 2016-10-27 10:40:31 +09:00
Ian Barwick
7ddb060bdc README: clarify that the repmgr metadatabase must be part of the replication cluster 2016-10-26 20:29:30 +09:00
Ian Barwick
6b02faf37c Update HISTORY 2016-10-24 09:04:04 +09:00
Ian Barwick
0cde0068dd repmgrd: improve witness failover logging 
Log the new master node as INFO rather than DEBUG.
 2016-10-20 10:56:16 +09:00
Ian Barwick
20d66df0ef repmgr: prevent involuntary cloning where no repmgr schema present on master 2016-10-19 20:26:55 +09:00
Ian Barwick
3f7c30b84d repmgr: require a valid repmgr cluster name unless -F/--force supplied 
Addresses issue mentioned in GitHub #242.
 2016-10-19 16:16:12 +09:00
Ian Barwick
a63baf7fcb repmgr: rename local variable config_file_found to remote_config_file_found 
We also have a global `config_file_found` - avoid confusion with that.
 2016-10-19 15:30:02 +09:00
Ian Barwick
e19c643389 repmgr: refactor check_upstream_config() for clarity 
Also ensure the check for sufficient walsenders is only carried out
when using pg_basebackup.
 2016-10-19 15:06:21 +09:00
Ian Barwick
f058833451 repmgr: in Barman clone mode, don't try and create the data directory twice 
In Barman mode the data directory is created early containing a temporary
directory needed to hold temporary files while cloning from the Barman
server. In other modes we might not know the data directory location until
connecting to the source server, so its creation happens later. In Barman
mode ensure that step is skipped.
 2016-10-19 10:00:50 +09:00
Ian Barwick
96c14adfdb repmgr: ensure data directory defaults to that of the source node 
As long as -D/--pgdata is not supplied, and barman mode is not used
for cloning.
 2016-10-17 16:34:49 +09:00
Ian Barwick
50119056a5 repmgr: clean up runtime options structure 
Place elements in a sensible order and split the associated initializer
macro over multiple lines for easier editing.

Also move a few related global variables into to the structure to keep
everything in the same place.
 2016-10-11 12:54:56 +09:00
Ian Barwick
a279c42df9 "pg_logical/snapshot" -> "pg_logical/snapshots" 2016-10-11 09:32:55 +09:00
Ian Barwick
f70b6ea136 Merge branch 'fix-barman-mode' of git://github.com/gciolli/repmgr 
Fixes GitHub #245.
 2016-10-11 09:21:31 +09:00
Ian Barwick
e4cb6d7130 repmgr: simplify LSN parsing 
Also silences a compiler warning.
 2016-10-10 22:56:49 +09:00
Gianni Ciolli
502c056753 When restoring from Barman, create pg_logical subdirectories 
Nodes cloned from Barman backups were missing these subdirectories,
and so they were unable to start when promoted.
 2016-10-10 10:18:33 +02:00
Ian Barwick
871ec47ff5 Fix repmgr cluster crosscheck output 
Show actual node ID rather than incremental number.

Fixes GitHub #244.
 2016-10-10 16:20:17 +09:00
Ian Barwick
f435abb3ec Update README 
Also reorder HISTORY entries.
 2016-10-10 15:10:06 +09:00
Ian Barwick
a217b4d0a9 repmgr: standardize SSH-related error messages 2016-10-07 07:42:15 +09:00
Ian Barwick
2dcb75f889 Add 'cluster crosscheck' to help output detail 
Per GitHub #243.
 2016-10-06 07:38:45 +09:00
Ian Barwick
b509ce6382 Minor README fix 2016-10-05 16:47:37 +09:00
Ian Barwick
1150bf272a Update README 
`--ignore-external-config-files` deprecated
 2016-10-05 15:09:07 +09:00
Ian Barwick
09ac6cd145 Update history 2016-10-05 13:57:10 +09:00
Ian Barwick
2fae788bc4 Add documentation for repmgrd failover process and failed node fencing 
Addresses GitHub #200.
 2016-10-05 11:25:36 +09:00
Ian Barwick
eb90f864c9 repmgr: consistent error message style 2016-10-05 10:31:25 +09:00
Ian Barwick
ba89758366 Update barman-wal-restore documentation 
Barman 2.0 provides this in a separate, more convenient `barman-cli` package;
document this and add note about previous `barman-wal-restore.py` script.
 2016-10-03 15:59:03 +09:00
Ian Barwick
84595fe711 Tweak repmgr.conf.sample 
Put `monitor_interval_secs` at the start of the `repmgrd` section, as it's
a very fundamental configuration item.
 2016-10-03 15:57:33 +09:00
Ian Barwick
9523894808 Bump dev version number 
3.3dev
 2016-09-30 15:14:35 +09:00
Ian Barwick
df09af4d57 Update README 
`repmgr cluster show --csv` now only reflects node availability,
and no longer overloads this with node type (master/standby) information.
 2016-09-30 15:07:34 +09:00
Ian Barwick
2c1cbc6bf9 Fix witness server initialisation 2016-09-30 13:43:47 +09:00
Ian Barwick
ed22fe326e Document and expand pg_ctl override configuration options 
These are now prefixed with "service_" to emphasize that they're
OS-level commands, not repmgr ones; also added reload and promote
commands:

    service_start_command
    service_stop_command
    service_restart_command
    service_reload_command
    service_promote_command

GitHub #169
 2016-09-30 11:58:45 +09:00
Ian Barwick
46500e1408 Documentation update and miscellaneous code cleanup 2016-09-30 09:30:22 +09:00
Ian Barwick
c3971513b6 Refactor show diagnose to handle node IDs correctly. 
See previous comments for `show matrix`.
 2016-09-30 01:46:01 +09:00
Ian Barwick
a2910eded9 Refactor show matrix to handle node IDs correctly. 
Previously the code assumed repmgr node IDs to be sequential,
which is not guaranteed to be the case. With a non-sequential
list of node IDs, an incorrect node id would be displayed,
and memory accessed beyond the bounds of the matrix array.

The refactored code is considerably less elegant than the original
but will correctly handle a non-sequential sequence of node IDs.
 2016-09-29 18:51:55 +09:00
Ian Barwick
dc70e2d804 Remove superfluous PQfinish() call 
Connection was previously closed, if this error condition is triggered
a segfault will occur.
 2016-09-29 11:53:13 +09:00