Daemon <daemon-name> timed out messages in ConfD

[arun88]

What happens to notification context when ConfD exits or aborts

The ConfD user guide says an application should use the confd_register_notification_stream() API to register with ConfD for notification streams and that the notification context (struct) is filled in by ConfD when performing this registration. Could you help me understand what happens when ConfD aborts or gracefully exits? the application really has a pointer to this notification context structure and it should not call confd_notification_send() with that notification context struct as it coul…

@per, thanks for the detailed response above. Starting a separate thread here. Just one more question.

“Note though that all notification contexts for a daemon are released by a call to confd_release_daemon() - i.e. after that call, all pointers to notifications contexts are invalid, and a call of confd_notification_send() with such a pointer is likely to result in a crash.”

When a daemon registers with ConfD using confd_init_daemon() call, under what circumstances would ConfD print log messages as follows in its log file:

“Daemon daemon-name timed out.”

I ask this because I am chasing an issue wherein, out of the blue, I started to see such messages in the ConfD log for my daemon and I would like to know where to start looking. These messages initially seem to come out every 20 minutes and the daemon eventually dies. I am not sure if the daemon died because ConfD had died minutes before. And I do not know if ConfD exits gracefully if such a daemon dies.

[per]

@arun88

under what circumstances would ConfD print log messages as follows in its log file:

“Daemon daemon-name timed out.”

Hm, I thought there was a mapping from those messages to “explanatory text” somewhere, but it seems the latter is only available for the “log numbers” that you would receive when subscribing to CONFD_NOTIF_DAEMON events - from $CONFD_DIR/include/confd_logsyms.h:

/* An external database daemon did not respond to a query.  */
#define  CONFD_DAEMON_TIMEOUT 15

Anyway, you will get more detailed info about the problem in the developerLog - if you set developerLogLevel to “trace”, you should also be able to figure out which particular request that timed out.

The timeouts can be configured via /confdConfig/capi/newSessionTimeout and /confdConfig/capi/queryTimeout - if either of them fires, ConfD will close all connections to the daemon and remove all its registrations.

@arun88

And I do not know if ConfD exits gracefully if such a daemon dies.

ConfD doesn’t exit at all if such a daemon dies, only logs it - in confd.log it will be (surprise:-):

Daemon <daemon-name> died