Currently it is not possible to get aborts due to for example SIGABRT to set an exit code != 0 which makes it hard for systemd to restart on failure. Either it is possible modify the current daemon implementation to handle this by using file descriptors etc but this is probably not a good idea since it might occur further problems down the line. A better approach is probably to change to use the stock python daemon which also is used by the other daemons like bdbserver and baltrad-exchange.
Currently it is not possible to get aborts due to for example SIGABRT to set an exit code != 0 which makes it hard for systemd to restart on failure. Either it is possible modify the current daemon implementation to handle this by using file descriptors etc but this is probably not a good idea since it might occur further problems down the line. A better approach is probably to change to use the stock python daemon which also is used by the other daemons like bdbserver and baltrad-exchange.