Struct Monitor 

pub struct Monitor { /* private fields */ }

The live monitor.

Implementations

impl Monitor

pub async fn bind(cfg: &Config) -> Result<Self>

Open the listening socket, if this arrangement needs one.

A failure to bind is fatal, exactly as in autossh: without the listener the loop can never complete and every probe would fail.

pub fn enabled(&self) -> bool

pub fn listener_fd(&self) -> Option<RawFd>

The descriptor the monitor listens on, if it has one.

Exposed so a test can assert it is close-on-exec: were the forked ssh to inherit it, the port or socket would stay held after rash exited.

pub fn next_forwards(&self) -> Vec<OsString>

The forwards for the next ssh start.

For the UNIX arrangement the remote socket path is different every time. A socket left behind on the remote by an unclean disconnect would stop sshd binding it, and StreamLocalBindUnlink defaults to no on the server — which a client cannot override. Without a fresh path, rash would restart for ever against a forward that could never come up.

pub async fn probe(&self, cfg: &Config) -> bool

Send a probe and wait for it to come back. false means the tunnel is not carrying traffic and ssh should be restarted.

Trait Implementations

impl Drop for Monitor

fn drop(&mut self)

Take the sockets away on the way out.

Binding already unlinks a stale file, so a leftover cannot break the next run — but without this the directory accumulates a dead pair per run, indefinitely. The pid file has the same guard, and the same limitation: a process killed with SIGKILL runs no destructors, so the next bind’s unlink is still the thing that guarantees correctness.

fn pin_drop(self: Pin<&mut Self>)

🔬This is a nightly-only experimental API. (pin_ergonomics)

Execute the destructor for this type, but different to Drop::drop, it requires self to be pinned.