github/repmgr
1
0
Fork 0
mirror of https://github.com/EnterpriseDB/repmgr.git synced 2026-03-26 00:26:30 +00:00
Code Issues Packages Projects Releases 1 Wiki Activity

repmgrd: during failover, check if a node was already promoted

Browse Source 
Previously, repmgrd assumed that during a failover, there would not
already be another primary node. However it's possible a node was
promoted manually. While this is not a desirable situation, it's
conceivable this could happen in the wild, so we should check for
it and react accordingly.

Also sanity-check that the follow target can actually be followed.

Addresses issue raised in GitHub #420.
This commit is contained in:
Ian Barwick
2019-03-21 16:16:59 +09:00
parent 7792de3543
commit 6441db23ff
4 changed files with 290 additions and 17 deletions
Download Patch File Download Diff File Expand all files Collapse all files

4
repmgr-client.c
View File

@@ -3144,6 +3144,8 @@ drop_replication_slot_if_exists(PGconn *conn, int node_id, char *slot_name)
/*
* Here we'll perform some timeline sanity checks to ensure the follow target
* can actually be followed.
*
* See also comment for check_node_can_follow() in repmgrd-physical.c .
*/
bool
check_node_can_attach(TimeLineID local_tli, XLogRecPtr local_xlogpos, PGconn *follow_target_conn, t_node_info *follow_target_node_record, bool is_rejoin)
@@ -3234,6 +3236,7 @@ check_node_can_attach(TimeLineID local_tli, XLogRecPtr local_xlogpos, PGconn *fo
return false;
}
/* timelines are the same - check relative positions */
if (follow_target_identification.timeline == local_tli)
{
XLogRecPtr follow_target_xlogpos = get_node_current_lsn(follow_target_conn);
@@ -3245,7 +3248,6 @@ check_node_can_attach(TimeLineID local_tli, XLogRecPtr local_xlogpos, PGconn *fo
return false;
}
/* timeline is the same - check relative positions */
if (local_xlogpos <= follow_target_xlogpos)
{
log_info(_("timelines are same, this server is not ahead"));