Closed InterLinked1 closed 6 days ago
I've seen several instances where an entire BBS server shuts down (cleanly) because a child process exited, but this shouldn't be happening.
e.g.:
[2023-08-17 12:30:45.829] VERBOSE[826200]: menu.c:606 bbs_menu_run: Node 4 selected option 'F' at menu(2) 'games' [2023-08-17 12:30:45.831] DEBUG[826200]: variables.c:496 bbs_node_substitute_vars: Substituted '/usr/games/go-fish' to '/usr/games/go-fish' [2023-08-17 12:30:45.831] DEBUG[826200]: handler.c:176 menu_handler_exec: Executing menu handler exec (/usr/games/go-fish) [2023-08-17 12:30:45.831] DEBUG[826200]: term.c:116 bbs_node_set_input: Node 4 (fd 103): input now buffered, echo enabled [2023-08-17 12:30:45.831] DEBUG[826200]: system.c:646 __bbs_execvpe_fd: node: 0x7fdb7c00f700, usenode: 1, fdin: -1, fdout: -1, filename: /usr/games/go-fish, isolated: no [2023-08-17 12:30:45.831] DEBUG[826200]: system.c:665 __bbs_execvpe_fd: sid: 56685, tcpgrp: -1 [2023-08-17 12:30:45.834] DEBUG[826200]: node.c:876 bbs_node_update_winsize: Sending SIGWINCH to foreground process 826222 for node 4 [2023-08-17 12:30:45.837] DEBUG[826200]: system.c:931 __bbs_execvpe_fd: Waiting for process 826222 to exit [2023-08-17 12:30:51.994] DEBUG[826200]: system.c:180 waitpidexit: Process 826222 (/usr/games/go-fish) killed, signal 2 [2023-08-17 12:30:51.994] DEBUG[826200]: system.c:197 waitpidexit: Command execution finished (/usr/games/go-fish): res = 0 [2023-08-17 12:30:51.994] DEBUG[826200]: socket.c:2368 bbs_node_wait_key: Waiting 120000 ms for any input [2023-08-17 12:30:51.994] DEBUG[826200]: term.c:116 bbs_node_set_input: Node 4 (fd 103): input now unbuffered, echo disabled [2023-08-17 12:30:51.994] DEBUG[826201]: pty.c:659 pty_master: Sending SIGINT to process 826222 [2023-08-17 12:30:51.995] DEBUG[779131]: bbs.c:620 __sigint_handler: Got SIGINT, requesting shutdown [2023-08-17 12:30:51.995] DEBUG[779131]: bbs.c:718 monitor_sig_flags: Shutdown requested [2023-08-17 12:30:51.995] DEBUG[779131]: event.c:127 bbs_event_broadcast: Event SHUTDOWN dispatched and not consumed [2023-08-17 12:30:51.995] VERBOSE[779131]: bbs.c:554 bbs_shutdown: Shutting down BBS [2023-08-17 12:30:51.995] DEBUG[779131]: node.c:489 node_shutdown: Terminating node 1 [2023-08-17 12:30:51.995] DEBUG[779131]: node.c:552 node_shutdown: Waiting for node 1 to exit
Marking as stale, I have not observed this in a very long time.
I've seen several instances where an entire BBS server shuts down (cleanly) because a child process exited, but this shouldn't be happening.
e.g.: