github/repmgr
1
0
Fork 0
mirror of https://github.com/EnterpriseDB/repmgr.git synced 2026-03-27 17:06:29 +00:00
Code Issues Packages Projects Releases 1 Wiki Activity

"standby promote": add check for WAL replay status if replay is paused

Browse Source 
If WAL replay is paused but WAL is still pending replay, PostgreSQL will ignore
the promote request until WAL replay is unpaused. This may lead to the standby
being promoted at an unpredictable point in time outside of repmgr's
control. Moreover it may not be obvious that this is happening, or why, and
it will appear that an apparently successful promotion attempt has not
actually worked.

To prevent this from happening, repmgr will now refuse to promote the
standy if WAL replay is paused *and* WAL is still pending replay.

GitHub #540.
This commit is contained in:
Ian Barwick
2019-02-05 13:25:21 +09:00
parent d8048060a2
commit 701944c194
4 changed files with 218 additions and 14 deletions
Download Patch File Download Diff File Expand all files Collapse all files

34
repmgr-action-standby.c
View File

@@ -1974,7 +1974,7 @@ do_standby_promote(void)
{
log_error(_("STANDBY PROMOTE can only be executed on a standby node"));
PQfinish(conn);
exit(ERR_BAD_CONFIG);
exit(ERR_PROMOTION_FAIL);
}
else
{
@@ -1988,6 +1988,38 @@ do_standby_promote(void)
log_info(_("node is a standby"));
}
/*
* Executing "pg_ctl ... promote" when WAL replay is paused and
* WAL is pending replay will mean the standby will not promote
* until replay is resumed.
*
* As that could happen at any time outside repmgr's control, we
* need to avoid leaving a "ticking timebomb" which might cause
* an unexpected status change in the replication cluster.
*/
if (is_wal_replay_paused(conn, true) == true)
{
ReplInfo replication_info;
init_replication_info(&replication_info);
log_error(_("WAL replay is paused on this node but not all WAL has been replayed"));
if (get_replication_info(conn, &replication_info) == true)
{
log_detail(_("replay paused at %X/%X; last WAL received is %X/%X"),
format_lsn(replication_info.last_wal_replay_lsn),
format_lsn(replication_info.last_wal_receive_lsn));
}
if (PQserverVersion(conn) >= 100000)
log_hint(_("execute \"pg_wal_replay_resume()\" to unpause WAL replay"));
else
log_hint(_("execute \"pg_xlog_replay_resume()\" to unpause WAL replay"));
PQfinish(conn);
exit(ERR_PROMOTION_FAIL);
}
/* check that there's no existing primary */
current_primary_conn = get_primary_connection_quiet(conn, &existing_primary_id, NULL);