ham_action_notify_pulse(), ham_action_notify_pulse_node()

Add a notify-pulse action to a condition


#include <ha/ham.h>

ham_action_t *ham_action_notify_pulse(
                 ham_condition_t *chdl,
                 const char *aname, 
                 int nd, 
                 pid_t topid,
                 int chid, 
                 int pulsecode, 
                 int value, 
                 unsigned flags);

ham_action_t *ham_action_notify_pulse_node(
                 ham_condition_t *chdl,
                 const char *aname, 
                 const char *nodename, 
                 pid_t topid,
                 int chid, 
                 int pulsecode, 
                 int value, 
                 unsigned flags);




These functions add an action to a given condition. This action will deliver a pulse notification to the process given in topid.

The handle (chdl) is obtained either:

The action is executed when the appropriate condition is triggered.

The nd specifies the node identifier of the remote node (or local node) to which the pulse is targeted The nd specified is the node identifier of the remote node at the time the call is made.

Note: Since node identifiers are transient objects, you should obtain the value for nd immediately before the call, using netmgr_strtond() or another function that converts nodenames into node identifiers.

Use the ham_action_notify_pulse_node() function when a nodename is used to specify a remote HAM instead of a node identifier (nd). The ham_action_notify_pulse_node() function takes as a parameter a fully qualified node name (FQNN)

The pulse in pulsecode with the given value will be delivered to topid on the specified channel ID (chid).

Currently the following flag is defined:

Indicates that the action is to be automatically rearmed after the entity that it belongs to is restarted. By default, this flag is disabled — actions automatically get pruned across restarts of the entity. Note that if the condition that this action belongs to is pruned after a restart, this action will also be removed, regardless of the value of this flag.
Indicates that if this action were to fail, and it is part of a list of actions, none of the actions following this one in the list of actions will be executed.
Indicates that the action will be retained even if it fails. The default behavior is to remove failed actions. Nevertheless if the condition that this action is associated with is not retained, the action will get automatically removed.

Users can specify what will be done if an action fails. By adding action fail actions to a list associated with an action, users can determine what will be done in the case of an action failure. For every action that fails, the corresponding action fail list will be executed. Certain actions (e.g. ham_action_log() and ham_action_heartbeat_healthy()) never fail. For more details, refer to the appropriate section in the HAM API reference for the ham_action_fail_*() calls.


A valid ham_action_t or NULL if an error occurred (errno is set).


Couldn't connect to the HAM.
The name provided for the action already exists.
The name given in aname is invalid (e.g. it contains the / character) or is NULL.

The connection to the HAM is invalid. This happens when the process that opened the connection (using ham_connect()) and the process that's calling ham_action_notify_pulse() aren't the same.

The name given is too long, i.e. it exceeds _POSIX_PATH_MAX (defined in <limits.h>). Note that the combined length of an entity/condition/action name is also limited by _POSIX_PATH_MAX.
There's no action of the name aname in a condition cname defined in an entity ename in the HAM's current context.
Not enough memory to create a new handle.

In addition to the above errors, the HAM returns any error it encounters while servicing this request.


QNX Neutrino

Cancellation point No
Interrupt handler No
Signal handler No
Thread Yes

See also:

ham_action_execute(), ham_action_notify_signal(), ham_action_remove(), ham_action_restart(), ham_action_waitfor(), ham_condition()