Fix stale master detection in multimaster mode
The MySQL monitor now correctly assigns stale status to master servers.
This commit is contained in:
parent
506ef1b9f6
commit
4cd36161ee
@ -1179,8 +1179,9 @@ void find_graph_cycles(MYSQL_MONITOR *handle, MONITOR_SERVERS *database, int nse
|
||||
monitor_clear_pending_status(graph[i].db, SERVER_SLAVE);
|
||||
}
|
||||
}
|
||||
else if (handle->detectStaleMaster && cycle == 0 && graph[i].cycle == 0 &&
|
||||
graph[i].db->server->status & SERVER_MASTER)
|
||||
else if (handle->detectStaleMaster && cycle == 0 &&
|
||||
graph[i].db->server->status & SERVER_MASTER &&
|
||||
(graph[i].db->pending_status & SERVER_MASTER) == 0)
|
||||
{
|
||||
/**
|
||||
* Stale master detection is handled here for multi-master mode.
|
||||
|
Loading…
x
Reference in New Issue
Block a user