I have to temporary NACK this one, because it needs a cman counter part
for all subsystems.
What problem are we effectively trying to solve here in the first place?
We had to opt for the reload_db option because we needed full
notification for any change happening in the objdb, instead of
installing several trackers.
Fabio
On 11/19/2012 2:34 PM, Jan Friesse wrote:
> confdb reload is used only for reloading of file, not for triggering
> changes.
>
> Signed-off-by: Jan Friesse <jfriesse@xxxxxxxxxx>
> ---
> tools/corosync-objctl.c | 15 ---------------
> 1 files changed, 0 insertions(+), 15 deletions(-)
>
> diff --git a/tools/corosync-objctl.c b/tools/corosync-objctl.c
> index 3e2bd4c..1168efe 100644
> --- a/tools/corosync-objctl.c
> +++ b/tools/corosync-objctl.c
> @@ -244,18 +244,6 @@ static void print_config_tree(confdb_handle_t handle, hdb_handle_t parent_object
> }
> }
>
> -static cs_error_t reload_db(confdb_handle_t handle)
> -{
> - char error[1024];
> - cs_error_t ret;
> -
> - ret = confdb_reload(handle, 0, error, sizeof(error) - 1);
> - if (ret != CONFDB_OK) {
> - fprintf (stderr, "Error reloading DB %d\n", ret);
> - }
> - return ret;
> -}
> -
> static int read_in_config_file (char * filename)
> {
> confdb_handle_t handle;
> @@ -326,7 +314,6 @@ static int read_in_config_file (char * filename)
> /* write the attribute */
> write_key (handle, line);
> }
> - reload_db(handle);
> confdb_finalize (handle);
> fclose (fh);
> return 0;
> @@ -931,8 +918,6 @@ int main (int argc, char *argv[]) {
> if (action == ACTION_TRACK) {
> listen_for_object_changes(handle);
> stop_tracking(handle);
> - } else {
> - reload_db(handle);
> }
>
> result = confdb_finalize (handle);
>
_______________________________________________
discuss mailing list
discuss@xxxxxxxxxxxx
http://lists.corosync.org/mailman/listinfo/discuss
