github/repmgr
1
0
Fork 0
mirror of https://github.com/EnterpriseDB/repmgr.git synced 2026-03-22 22:56:29 +00:00
Code Issues Packages Projects Releases 1 Wiki Activity
1,197 Commits 43 Branches 68 Tags
48a2274b113ffe5f323710140121c8746ba5b596
Commit Graph

1197 Commits

This Branch
This Branch
All Branches
Author SHA1 Message Date
Ian Barwick
48a2274b11 Use appendPQExpBufferStr where approrpriate 2019-03-06 13:23:38 +09:00
Ian Barwick
19bcfa7264 Rename "..._primary_last_seen" functions to "..._upstream_last_seen" 
As that better reflects what they do.
 2019-03-06 13:23:33 +09:00
Ian Barwick
486877c3d5 repmgrd: log details of nodes which can see primary 
If a failover is cancelled because other nodes can still see the primary,
log the identies of those nodes.
 2019-03-06 13:23:27 +09:00
Ian Barwick
9753bcc8c3 repmgrd: during failover, check if other nodes have seen the primary 
In a situation where only some standbys are cut off from the primary,
a failover would result in a split brain/split cluster situation,
as it's likely one of the cut-off standbys will promote itself, and
other cut-off standbys (but not all standbys) will follow it.

To prevent this happening, interrogate the other sibiling nodes to
check whether they've seen the primary within a reasonably short interval;
if this is the case, do not take any failover action.

This feature is experimental.
 2019-03-06 13:23:22 +09:00
Ian Barwick
bd35b450da daemon status: with csv output, show repmgrd status as unknown where appropriate 
Previously, if PostgreSQL was not running on the node, repmgrd and
pause status were shown as "0", implying their status was known.

This brings the csv output in line with the human-readable output,
which displays "n/a" in this case.
 2019-02-28 12:28:04 +09:00
Ian Barwick
1f256d4d73 doc: upate release notes 2019-02-28 10:02:05 +09:00
Ian Barwick
1524e2449f Split command execution functions into separate library 
These may need to be executed by repmgrd.
 2019-02-27 14:41:38 +09:00
Ian Barwick
0cd2bd2e91 repmgrd: add additional logging during a failover operation 2019-02-27 11:45:34 +09:00
Ian Barwick
98b78df16c Remove unneeded debugging output 2019-02-26 21:17:17 +09:00
Ian Barwick
b946dce2f0 doc: update introductory blurb 2019-02-26 15:19:41 +09:00
Ian Barwick
39234afcbf standby clone: check upstream connections after data copy operation 
With long-running copy operations, it's possible the connection(s) to
the primary/source server may go away for some reason, so recheck
their availability before attempting to reuse.
 2019-02-26 14:37:51 +09:00
John Naylor
23569a19b1 Doc fix: PostgreSQL 9.4 is no longer considered recent 2019-02-25 13:02:56 +09:00
John Naylor
c650fd3412 Fix typo 2019-02-25 13:02:51 +09:00
Ian Barwick
c30e65b3f2 Add some missing query error logging 2019-02-25 13:02:45 +09:00
Ian Barwick
07097575b1 daemon status: add column "upstream last seen" 
This displays the interval (in seconds) since the repmgrd instance on
each node last confirmed its upstream node is available.
 2019-02-23 13:03:16 +09:00
Ian Barwick
71d151ca87 Don't check status of logical replication slots 
We only want to check the status of physical replication slots
to determine whether a streaming replication standby has become
detached and there is therefore a risk of uncontrolled WAL buildup
on the local node.

It's not feasible to second-guess the state of logical replication
slots.
 2019-02-23 10:09:43 +09:00
Ian Barwick
5abec2bb97 doc: clarify replication slot usage with Barman 
Barman will usually use one replication slot, but that's generally
preferable to multiple slots.
 2019-02-22 13:52:02 +09:00
Ian Barwick
de70fd42dc node check: simplify output generation in --is-shutdown-cleanly check 2019-02-22 10:49:06 +09:00
Ian Barwick
99550b91bd standby register: warn if standby is running and connection params provided 
Addresses GitHub #552.
 2019-02-22 10:31:00 +09:00
John Naylor
70190c37c4 Bring list of supported versions on the doc front page in line with the supported version matrix 2019-02-20 11:41:17 +07:00
Ian Barwick
f3fc4e5afb Minor syntax formatting tweak 
For consistency.
 2019-02-15 19:58:35 +09:00
Ian Barwick
629c552348 primary unregister: ensure correct behaviour when executed on a witness 
Fixes GitHub #548.
 2019-02-15 19:49:17 +09:00
Ian Barwick
85a97c933f Handle unhandled NodeStatus in switch statement 2019-02-15 19:31:06 +09:00
Ian Barwick
3a5a4388c7 cluster show: differentiate unreachable status 
Differentiate between unreachable nodes and nodes which are running
but rejecting connections.
 2019-02-15 16:01:55 +09:00
Ian Barwick
9338a9e233 Improve logging output 
Avoid emitting blank detail lineImprove logging output

Avoid emitting blank detail lineImprove logging output

Avoid emitting blank detail lineImprove logging output

Avoid emitting blank detail lineImprove logging output

Avoid emitting blank detail lineImprove logging output

Avoid emitting blank detail lineImprove logging output

Avoid emitting blank detail lineImprove logging output

Avoid emitting blank detail lineImprove logging output

Avoid emitting blank detail line
 2019-02-15 10:49:56 +09:00
Ian Barwick
7fad2ed2c8 standby switchover: improve error output 
It wasn't clear why repmgr thinks the demotion candidate is not
the upstream of the promotion candidate.
 2019-02-14 17:22:24 +09:00
Ian Barwick
9305953bd2 Fix history file parsing 
Also add additional debugging output.
 2019-02-14 15:52:40 +09:00
Ian Barwick
aeb9639ed9 node rejoin: add more log detail during rejoin success check 
Stating what is actually being checked where might be useful
when diagnosing potential issues.
 2019-02-13 15:29:39 +09:00
Ian Barwick
bc9e725d05 node rejoin: always emit detail about relative LSNs 
Previously repmgr only emitted that if there was a timeline/LSN
mismatch, but it's useful to have confirmation of how it came
to the conclusion that rejoin will succeed.
 2019-02-13 15:16:40 +09:00
Ian Barwick
905e108f8f doc: fix typos etc. in "standby follow" reference 2019-02-12 17:24:56 +09:00
Ian Barwick
f2362a06fa doc: update "standby switchover" reference 2019-02-12 16:39:13 +09:00
Ian Barwick
7b85cb9f12 doc: update "standby follow" reference 
Add note about handling of timeline forks etc.
 2019-02-12 16:39:06 +09:00
Ian Barwick
790bec21dd node rejoin: handle case where node to rejoin was primary 
In that case the minRecoveryPoint* fields may be empty.
 2019-02-12 13:31:25 +09:00
Ian Barwick
a0dc673439 "node rejoin": use minRecoveryPointTLI for comparing timelines 2019-02-12 13:31:21 +09:00
Ian Barwick
25019d1cc5 Refactor is_wal_replay_paused() query 
Make sure it doesn't emit an error if executed on a node not
in recovery.

The caller should theoretically only execute it on nodes in
recovery, but there are sure to be corner cases where the node
has come out of recovery.
 2019-02-12 10:21:05 +09:00
Ian Barwick
d00cb767a6 cluster show: don't try to run WAL replay pause query on unreachable node 2019-02-12 10:15:06 +09:00
Ian Barwick
8e0d28d8dc Fix "repmgr daemon --help" output 
Per report from Shaun.
 2019-02-12 09:20:29 +09:00
yonj1e
e146fb4fc3 Fix undeclared 'TRUE' error 
GitHub #547.
 2019-02-11 16:55:54 +09:00
Ian Barwick
8773543e10 doc: update "daemon (start|stop)" documentation 
Clarify various aspects related to configuration.
 2019-02-11 10:55:10 +09:00
Ian Barwick
a4cd4ee553 doc: fix quoting in "standby switchover" index entries 2019-02-11 10:34:02 +09:00
Ian Barwick
a61dd8a750 doc: tweak support text 2019-02-08 15:28:12 +09:00
Ian Barwick
2c84716e66 doc: add information about reporting issues etc. 
Useful to have a linkable document listing the information required
to have a chance of troubleshooting issues.
 2019-02-08 11:55:42 +09:00
Ian Barwick
f1667a7e98 repmgrd: don't consider nodes where repmgrd is not running 
If, for whatever reason, repmgrd is not running on a node, but that
node qualifies as promotion candidate, failover will not take place
as that node will never promote itself.

We therefore discount nodes where repmgrd is running as promotion
candidates, which will ensure one node is always promoted.

There is a slight risk here that the node(s) where repmgrd is not running
are further ahead, leading to a timeline fork. It might be possible
to mitigate that by having the "election" leader perform the promote
(or follow) operation.
 2019-02-07 17:07:13 +09:00
Ian Barwick
b91900f831 doc: clarify "repmgr daemon status" CSV output 2019-02-07 14:55:42 +09:00
Ian Barwick
aa1e64ec11 Warn about redundant use of --compact option 2019-02-07 14:35:30 +09:00
Ian Barwick
5d6037303b "daemon status": display node priority 
GitHub #541.
 2019-02-07 14:35:24 +09:00
Ian Barwick
8aaf6571a0 "cluster show": display node priority 
GitHUb #541.
 2019-02-07 14:35:21 +09:00
Ian Barwick
9433f80364 "cluster show": warn about nodes with paused WAL replay 
We do this in "repmgr daemon status" already, so do it here too for consistency.

Related to GitHub #540.
 2019-02-07 13:48:46 +09:00
Ian Barwick
aee13aee52 doc: note repmgrd behaviour when WAL replay is paused 
Related to GitHub #540.
 2019-02-07 13:28:29 +09:00
Ian Barwick
f0a0be0248 Remove pointless default allocation in _get_node_record() 2019-02-07 11:41:08 +09:00