609453a848
Handle failover of top-level standby
...
Cascaded standbys will not go into failover so we need to ignore
these when looking for candidates for promotion.
2015-01-16 12:35:01 +09:00
a82d37e48a
Improve node metadata and upstream connecting mechanism
...
To handle cascaded replication we're going to have to keep track
of each node's upstream node. Also enumerate the node type
("primary", "standby" or "witness") and mark if active.
2015-01-16 10:28:02 +09:00
4b6c097a3e
Use AS for consistency
2015-01-16 08:46:28 +09:00
3dfa33d01d
Store upstream node ID if set
...
Required to manage cascaded standbys
2015-01-16 00:29:43 +09:00
a20afe28ec
Add upstream_node_id to repl_nodes table
2015-01-15 23:54:07 +09:00
692204e381
Consolidate duplicated schema check code
2015-01-15 18:34:51 +09:00
2ae27521a3
Some infrastructure for supporting cascading replication
...
Does not fully work yet.
2015-01-15 15:37:09 +09:00
d141d2a8aa
Remove unused variable
2015-01-14 16:25:24 +09:00
fc6f5ddaa7
Add function repmgr_get_primary_conninfo()
2015-01-14 16:23:08 +09:00
4b7fb70de3
Add current year to copyright
2015-01-14 15:50:42 +09:00
5db6489ae5
Initial support for cascaded standbys
2015-01-14 14:18:57 +09:00
96f6695a52
Clean up repmgr_get_last_standby_location() polling loop
...
No need to reconnect on every iteration of the loop. Move conditions
which break out of the loop straight away out of the loop.
2015-01-13 16:08:36 +09:00
f006a79af9
Clean up log messages for invalid LSNs
2015-01-13 14:53:55 +09:00
3a8aa86e0c
Remove pointless check
2015-01-13 14:25:58 +09:00
9cbac64ad4
Remove pointless check
...
Witness server can never be considered as a promotion candidate.
2015-01-13 14:24:19 +09:00
242fc8416c
Consolidate disconnection calls
2015-01-13 13:59:27 +09:00
355d8b8f01
Use InvalidXLogRecPtr for pedantic correctness
2015-01-13 13:47:22 +09:00
8df7966540
Rename variable
...
"find_best" sounds like an imperative.
2015-01-13 13:42:30 +09:00
1a790549ea
Refactor if-statement
2015-01-13 13:40:47 +09:00
b80d25ad33
Add some hopefully correct comments for future reference
...
Also some debugging output which will need to be removed later.
2015-01-13 11:12:20 +09:00
a1a61e5ec3
Remove unneeded include
2015-01-13 07:57:54 +09:00
1221b73293
repmgrd: explicitly state INSERT statement target columns
2015-01-12 20:47:22 +09:00
93d5d7905f
Remove "XL*" macros and rationalize XLogRecPtr handling
...
As of 9.3, XLogRecPtr is a uint64, which makes it much easier to
handle, so we can retire the legacy macros and work directly
with the values.
2015-01-12 17:18:57 +09:00
41a5274b44
Comment for consistency
2015-01-12 15:30:17 +09:00
d9f83cf620
Replace hard-coded value with local_options.reconnect_attempts;
2015-01-12 14:09:44 +09:00
905a2d0a5e
Have witness server retry determining the new master
...
Loop a few times before giving up
2015-01-12 14:05:37 +09:00
526803c5e0
Witness server should always try to re-connect to primary
...
Standbys will try to reconnect even in MANUAL_FAILOVER mode,
so it's consistent to have the witness server do that too.
2015-01-12 13:32:04 +09:00
2ff24fecf2
Remove redundant constant
2015-01-12 13:16:13 +09:00
a5ead16d6f
Change query to use COUNT(*)
...
No point implying we're fetching the entire record
2015-01-12 12:28:06 +09:00
5fb84b9627
Witness server: on failover attempt to reconnect to new master
...
Previously it was just quitting.
2015-01-12 11:23:20 +09:00
437485bf6a
Fix debugging output
2015-01-12 09:47:13 +09:00
822867ec24
Fix thinko
2015-01-12 09:40:47 +09:00
5b88a980b8
Add some debugging output
2015-01-12 07:57:42 +09:00
e958b8f2d7
Fix typo
2015-01-09 18:44:31 +09:00
74a963a10e
Fix schema quoting
...
There was a lot of duplicated/unused related to handling the
schema name; consolidated and rationalised.
2015-01-09 15:51:34 +09:00
dd064cb47c
res1 -> res
2015-01-09 11:20:07 +09:00
17de82493a
Consolidate repl_node insert statements
2015-01-09 11:14:47 +09:00
f2fa60f5cf
Tidy up SQL statements
...
Improves readability
2015-01-09 09:46:41 +09:00
acf2744ed7
Replace magic number with defined constant
...
And use it to restrict the number of rows retrieved to avoid
overflowing the array.
2015-01-08 17:19:02 +09:00
7bafd490c1
Remove #define for pre-9.3 versions
2015-01-08 16:59:10 +09:00
7c15176646
Clean up comments
2015-01-08 14:04:43 +09:00
97cbd7d557
Remove superfluous variable
2015-01-08 13:23:33 +09:00
5798241205
Add 'recovery_target_timeline = latest' to recovery.conf
...
Also revert to generating our own recovery.conf rather than using
pg_basebackup's minimal file.
2015-01-08 10:46:49 +09:00
5b5b4cd4cd
Remove redundant comment
2015-01-08 09:51:53 +09:00
9dd78f34db
Add missing short options to --help output
2015-01-07 08:56:05 +09:00
a5b5e3d384
No need to store master version number here
2015-01-07 08:55:17 +09:00
241622694d
Remove support for pre-9.3 servers
2015-01-06 14:53:26 +09:00
687872e979
get_data_directory() -> get_pg_setting()
...
More code consolidation
2015-01-06 13:47:31 +09:00
3033f2dfaf
Fix get_cluster_size()
...
Was returning a pointer to a cleared PQresult
2015-01-06 10:30:33 +09:00
718024454e
Add function get_data_directory()
...
Consolidate duplicate code
2015-01-06 10:06:58 +09:00
Ian Barwick