951c7dbd07
repmgrd: in BDR mode, have each repmgrd monitor each node
...
This will cover both the case when an entire node including
repmgrd goes down, and when one PostgreSQL instance goes down
but repmgrd is still up (in which case only one of the repmgrds
will handle the failover).
2017-07-14 15:01:18 +09:00
e3b3fb65f0
repmgrd: restrict BDR monitoring to two node setup
...
It's not safe to have more than two nodes with this kind of
"failover", so we don't need to select alternative nodes by
priority.
2017-07-14 12:56:11 +09:00
dfcf85a62f
repmgrd: further BDR sanity checks
2017-07-14 10:27:28 +09:00
7eadbf6b17
Various improvements to "repmgr bdr register/unregister"
2017-07-12 22:38:03 +09:00
0a1addfdc0
When registering a BDR node, sync repmgr.nodes from another node
...
If a BDR node is added via bdr_group_join(), repmgr.nodes will
start off empty, so we'll need to sync it ourselves before adding
it to the repmgr replication set.
2017-07-12 10:11:25 +09:00
1cccb1dd5a
Add "repmgr bdr unregister"
2017-07-12 10:11:21 +09:00
71a0871232
Add "repmgr bdr register"
2017-07-11 15:38:58 +09:00
2962ffe605
repmgrd: initial BDR monitoring support
2017-07-10 23:58:59 +09:00
dddea9814b
Add BDR-related database functions
2017-07-10 21:52:39 +09:00
5fbcf3e476
Remove witness server references
2017-07-10 09:31:31 +09:00
0d226867b4
Add "location" column
2017-07-06 01:17:00 +09:00
9351e532b4
Ensure configuration parameter "replication_user" is smaller than NAMEDATALEN
2017-07-06 00:22:23 +09:00
617dee6bd6
Add function create_event_record()
...
For logging an event to the event table without generating an external
event notification.
Rename existing create_event_record*() functions to create_event_notification*()
as this describes their function better.
2017-07-05 09:52:22 +09:00
24c6b2c9f1
repmgrd: initial code for cascaded standby failover
2017-07-04 23:14:05 +09:00
618a2346e1
repmgrd: various fixed, mainly clearing status after a failover event
2017-07-04 11:55:03 +09:00
debe5a18c5
have new primary communicate to standbys
2017-06-30 21:45:25 +09:00
fc4f276844
Improve handling
...
not sure if we need to store the electoral term...
2017-06-30 13:40:19 +09:00
3514e20367
poke it around until it works less badly
2017-06-29 09:35:09 +09:00
fa86fe4ad8
Basic voting
2017-06-29 01:11:21 +09:00
d6b6255144
interim commit
2017-06-28 18:20:03 +09:00
ded8d95e5a
interim commit
2017-06-28 16:38:41 +09:00
35b6178e07
placeholder code for function
2017-06-27 09:50:47 +09:00
78a16d746d
Initial primary node monitoring
2017-06-27 00:15:29 +09:00
46c956e61a
Use "primary" instead of "master"
2017-06-23 21:33:54 +09:00
1b2652037d
Rename enum types for consistency
2017-06-23 16:38:14 +09:00
dbaa2e0b44
Add a RecordStatus return type for functions which populate record structures
...
Unify a bunch of slightly different ways of handling the result.
2017-06-23 16:16:46 +09:00
6cdf73b4cb
repmgr standby promote: suppress master database connection error message
...
Otherwise the first line of output is an ERROR, which is confusing,
even though it's expected.
2017-06-21 13:21:44 +09:00
94a88326ef
repmgrd: further code ported
2017-06-20 09:17:29 +09:00
030fdc046b
repmgr standby follow: main code
2017-06-16 21:38:53 +09:00
36b3782009
Store the replication user in repmgr.nodes
...
When creating recovery.conf outside of "repmgr standby clone",
there was no way of knowing if a replication user had been
explicitly provided with --replication-user, meaning the value
of "primary_conninfo" would be set to the "conninfo" field of the
node's upstream node record.
We'll add an extra column to store the replication user for each
node so it can be referenced at any time.
2017-06-14 23:27:26 +09:00
e89c43c5cb
Remove unused backup functions
...
Not needed since removal of rsync functionality
2017-06-13 00:35:01 +09:00
bb7d3e41c3
repmgr master unregister: check for downstream nodes
...
Foreign key dependencies will make it impossible to remove the node
if it still has downstream nodes pointing to it.
2017-06-12 22:24:50 +09:00
aa53514f9f
repmgr: various fixes for "master unregister"
2017-06-12 08:18:10 +09:00
fef184ce9a
Initial implementation of "repmgr master unregister"
...
Also adds "--dry-run" option
2017-06-12 08:17:58 +09:00
124398bed5
Replace is_standby() with get_recovery_type()
...
We what to know what kind of node it is, not whether it's a standby or not.
2017-06-09 11:25:43 +09:00
51da33a260
Initial implementation of "standby promote"
2017-05-10 09:24:51 +09:00
c7f49541a3
Add "standby unregister"
2017-05-08 23:37:53 +09:00
7844afa074
Implement "standby register"
2017-05-05 10:50:41 +09:00
06b79f2263
"standby clone": ensure replication user is used when conninfo str provided
2017-05-05 00:13:04 +09:00
10959d98bc
"standby clone": honour -S/--superuser setting
...
Refactor superuser connection detection.
2017-05-02 17:58:57 +09:00
00fd8e81f8
"standby clone": basic port complete
2017-05-02 14:25:08 +09:00
a960ed9d49
"standby clone": perform clone operations
2017-05-02 09:58:17 +09:00
8f74d1b946
"standby clone": create replication slot
2017-05-02 00:38:36 +09:00
bcb17dd71a
More functionality for "standby clone"
2017-04-29 22:41:43 +09:00
dc347f1484
Additional "standby clone" code
...
We'll break up the unwieldy "do_standby_clone()" function into discrete
unit for easier maintenance.
2017-04-28 22:00:26 +09:00
cd0509194d
Validate --node-id/--node-name, if provided and acceptable
...
Also update --help output.
2017-04-26 21:18:28 +09:00
ef5c6b37a4
Add event logging
2017-04-24 23:20:23 +09:00
8a04879953
Parameterize record creation/update queries
...
Also merge the main functionality into a common function to avoid
duplication.
2017-04-24 22:21:38 +09:00
ca6b493763
create and update node records
2017-04-24 10:24:44 +09:00
18c2cfda0b
Fix tabulation
2017-04-22 09:06:42 +09:00
Ian Barwick