github/repmgr
1
0
Fork 0
mirror of https://github.com/EnterpriseDB/repmgr.git synced 2026-03-26 08:36:30 +00:00
Code Issues Packages Projects Releases 1 Wiki Activity

Improve handling of event logging in rempgrd

Browse Source 
Provide the master connection if available, and if not enable
create_event_record() to skip trying to write to the database,
but execute the notification program if defined.
This commit is contained in:
Ian Barwick
2015-03-24 13:40:39 +09:00
parent 2cadb3424d
commit bd19a2c868
2 changed files with 92 additions and 64 deletions
Download Patch File Download Diff File Expand all files Collapse all files

38
repmgrd.c
View File

@@ -410,12 +410,15 @@ main(int argc, char **argv)
_("unable to connect to master node '%s'"),
local_options.cluster_name);
create_event_record(my_local_conn,
log_err("%s\n", errmsg.data);
create_event_record(NULL,
&local_options,
local_options.node,
"repmgrd_shutdown",
false,
errmsg.data);
terminate(ERR_BAD_CONFIG);
}
@@ -586,8 +589,10 @@ witness_monitor(void)
appendPQExpBuffer(&errmsg,
_("unable to determine a valid master node, terminating..."));
log_err("%s\n", errmsg.data);
create_event_record(my_local_conn,
create_event_record(NULL,
&local_options,
local_options.node,
"repmgrd_shutdown",
@@ -659,7 +664,7 @@ witness_monitor(void)
/*
* Insert monitor info, this is basically the time and xlog replayed,
* applied on standby and current xlog location in primary.
* Also do the math to see how far are we in bytes for being uptodate
* Also do the math to see how far are we in bytes for being up-to-date
*/
static void
standby_monitor(void)
@@ -705,12 +710,13 @@ standby_monitor(void)
log_err("%s\n", errmsg.data);
create_event_record(my_local_conn,
create_event_record(primary_conn,
&local_options,
local_options.node,
"repmgrd_shutdown",
false,
errmsg.data);
terminate(ERR_DB_CON);
}
@@ -777,9 +783,10 @@ standby_monitor(void)
appendPQExpBuffer(&errmsg,
_("Unable to reconnect to master after %i attempts, terminating..."),
local_options.reconnect_attempts);
log_err("%s\n", errmsg.data);
create_event_record(my_local_conn,
create_event_record(NULL,
&local_options,
local_options.node,
"repmgrd_shutdown",
@@ -819,11 +826,11 @@ standby_monitor(void)
initPQExpBuffer(&errmsg);
appendPQExpBuffer(&errmsg,
_("unable to reconnect to new upstream node, terminating...")
);
_("unable to reconnect to new upstream node, terminating..."));
log_err("%s\n", errmsg.data);
create_event_record(my_local_conn,
create_event_record(primary_conn,
&local_options,
local_options.node,
"repmgrd_shutdown",
@@ -1434,7 +1441,9 @@ do_primary_failover(void)
node_info.node_id,
failed_primary.node_id);
create_event_record(my_local_conn,
log_err("%s\n", event_details.data);
create_event_record(NULL,
&local_options,
node_info.node_id,
"repmgrd_failover_promote",
@@ -1444,7 +1453,7 @@ do_primary_failover(void)
terminate(ERR_DB_QUERY);
}
/* update internal record for this node*/
/* update internal record for this node */
node_info = get_node_info(my_local_conn, local_options.cluster_name, local_options.node);
appendPQExpBuffer(&event_details,
@@ -1452,6 +1461,7 @@ do_primary_failover(void)
node_info.node_id,
failed_primary.node_id);
/* my_local_conn is now the master */
create_event_record(my_local_conn,
&local_options,
node_info.node_id,
@@ -1500,6 +1510,8 @@ do_primary_failover(void)
node_info.slot_name,
PQerrorMessage(new_primary_conn));
log_err("%s\n", event_details.data);
create_event_record(new_primary_conn,
&local_options,
node_info.node_id,
@@ -1507,8 +1519,6 @@ do_primary_failover(void)
false,
event_details.data);
log_err("%s\n", event_details.data);
PQfinish(new_primary_conn);
terminate(ERR_DB_QUERY);
}
@@ -1532,6 +1542,8 @@ do_primary_failover(void)
node_info.node_id,
best_candidate.node_id);
log_err("%s\n", event_details.data);
create_event_record(new_primary_conn,
&local_options,
node_info.node_id,
@@ -1685,7 +1697,7 @@ check_connection(PGconn *conn, const char *type)
int connection_retries;
/*
* Check if the master is still available if after
* Check if the node is still available if after
* local_options.reconnect_attempts * local_options.reconnect_intvl
* seconds of retries we cannot reconnect return false
*/