Expand description
cfg is the internal configuration and information library for corosync, it is mainly used by internal tools but may also contain API calls useful to some applications that need detailed information about or control of the operation of corosync and the cluster.
Structs§
- Callbacks
- Callback from track_start. Will be called if another process requests to shut down corosync. reply_to_shutdown should be called with a ShutdownReply of either Yes or No.
- Handle
- A handle into the cfg library. returned from initialize and needed for all other calls
- Link
Status - Status of a link inside NodeStatus struct
- Node
Status - Structure returned from node_status_get, shows all the details of a node that is known to corosync, including all configured links
Enums§
- Node
Status Version - Version of the NodeStatus structure returned from node_status_get
- Shutdown
Flags - Flags for try_shutdown
- Shutdown
Reply - Responses for reply_to_shutdown
- Track
Flags - Trackflags for track_start. None currently supported
Functions§
- dispatch
- Call any/all active CFG callbacks for this Handle see DispatchFlags for details
- fd_get
- Returns a file descriptor to use for poll/select on the CFG handle
- finalize
- Finish with a connection to corosync, after calling this the Handle is invalid
- initialize
- Initialize a connection to the cfg library. You must call this before doing anything else and use the passed back Handle. Remember to free the handle using finalize when finished.
- kill_
node - Tell another cluster node to shutdown. reason is a string that will be written to the system log files.
- local_
get - Get the local NodeId
- node_
status_ get - Get the extended status of a node in the cluster (including active links) from its NodeId. Returns a filled in NodeStatus struct
- reload_
cnfig - Reload the cluster configuration on all nodes
- reopen_
log_ files - Re-open the cluster log files, on this node only
- reply_
to_ shutdown - Reply to a shutdown request with Yes or No ShutdownReply
- track_
start - Start tracking for shutdown notifications
- track_
stop - Stop tracking for shutdown notifications
- try_
shutdown - Ask this cluster node to shutdown. If ShutdownFlags is set to Request then it may be refused by other applications that have registered for shutdown callbacks.