What causes "syslog is already initialized" in shadowsocks/shadowsocks-rust?

Initialize logging/syslog exactly once per process; reuse the existing writer

syslog is already initialized

Error message

syslog is already initialized

What it means

The syslog crate's Syslog::new returns None when a syslog connection was already opened in the process (only one syslog target per process). make_syslog_writer panics in that case because it cannot create a second writer. This happens when syslog logging is initialized more than once, e.g. building the logger layer twice.

Source

Thrown at src/logging/tracing.rs:217

            9 => Facility::Cron,
            10 => Facility::AuthPriv,
            16 => Facility::Local0,
            17 => Facility::Local1,
            18 => Facility::Local2,
            19 => Facility::Local3,
            20 => Facility::Local4,
            21 => Facility::Local5,
            22 => Facility::Local6,
            23 => Facility::Local7,
            _ => panic!("unsupported syslog facility: {}", f),
        },
    };
    let options = Options::default();
    let identity = CString::new(identity).expect("syslog identity contains null-byte ('\\0')");

    match Syslog::new(identity, options, facility) {
        Some(l) => l,
        None => panic!("syslog is already initialized"),
    }
}

View on GitHub (pinned to 8eb0f0a65b)

Solutions

Initialize logging/syslog exactly once per process; reuse the existing writer
Guard logger setup with std::sync::Once or a OnceLock
If reconfiguring, tear down and don't call Syslog::new again — the connection persists

Example fix

// before
fn init_logs() {
    make_layer(); // called twice -> panic
}
// after
static INIT: Once = Once::new();
fn init_logs() {
    INIT.call_once(make_layer);
}

Defensive patterns

Strategy: validation

Validate before calling

static SYSLOG_INIT: std::sync::Once = std::sync::Once::new();
SYSLOG_INIT.call_once(|| { make_layer(); }); // prevent second Syslog::new

Prevention

Initialize logging once per process using Once/OnceLock
Do not rebuild logging layers at runtime if syslog is among them
In tests, use one shared syslog init or skip syslog targets
Document that only one syslog writer may exist per process

When it happens

Trigger: Calling make_syslog_writer/make_layer a second time after a prior Syslog::new succeeded — e.g. re-running logging initialization, or initializing both a local logger and an additional syslog layer that each call Syslog::new.

Common situations: Calling initialize_logging twice during startup; reconfiguring logging at runtime by rebuilding layers; tests that initialize syslog logging repeatedly in one process.

Understand the failure class

Background: "Invalid state transition" errors: "status must be X, actually Y", "already rejected/charging/uninstalled", "cannot ... while running" — what they mean when a library rejects your call — this error's family across 31 libraries.

Related errors

AI-assisted analysis of shadowsocks/shadowsocks-rust@8eb0f0a65b (2026-09-09). Data as JSON: /api/errors/bb6d1e98ad4feb3a. Report an issue: GitHub.