BoundaryML/baml · error · BridgeError

call_id {0} is already in use by an active call

Error message

call_id {0} is already in use by an active call

What it means

BridgeError::DuplicateCallId carries the conflicting id. The bridge tracks in-flight calls by call_id so results can be routed back to the right caller; registering a new call with an id that already belongs to an active call is rejected to prevent result cross-wiring.

Source

Thrown at baml_language/crates/bridge_cffi/src/error.rs:37

    #[error("{0}")]
    Runtime(#[from] bex_project::RuntimeError),

    #[error("CallFunctionArgs.call_target must be set")]
    MissingCallTarget,

    #[error("type arguments are not supported when invoking a BAML function handle")]
    FunctionHandleTypeArgs,

    #[error("Function not found: {name}")]
    FunctionNotFound { name: String },

    #[error("Missing argument '{parameter}' for function '{function}'")]
    MissingArgument { function: String, parameter: String },

    #[error("Not implemented: {0}")]
    NotImplemented(String),

    #[error("call_id {0} is already in use by an active call")]
    DuplicateCallId(u64),

    #[error("call_id must be a nonzero uint64")]
    InvalidCallId,

    #[error("Internal error: {0}")]
    Internal(String),

    #[error("{0}")]
    Startup(String),
}

View on GitHub (pinned to bd85ce9dee)

Solutions

1. Generate unique call_ids (e.g. monotonic counter or UUID-derived integer) for every call.
2. Wait for or cancel the active call holding that id before reusing it.
3. Fix id generators that reset between retries or threads.
4. Consume pending results promptly so ids retire.

Example fix

// before
let call_id = 42; // reused every call
// after
static NEXT_ID: AtomicU64 = AtomicU64::new(1);
let call_id = NEXT_ID.fetch_add(1, Ordering::Relaxed);

if call_id in active_calls:
    raise ValueError(f'call_id {call_id} still active')

try:
    bridge.call_start(args)
except BridgeError as e:
    if e is BridgeError.DuplicateCallId:
        args.call_id = new_unique_id()
        bridge.call_start(args)

When it happens

Trigger: Starting a new asynchronous bridge call with a call_id equal to one from a call that has not yet completed or been consumed; reusing ids without waiting for completion; calling code that resets its id counter while calls are still active.

Common situations: Retry logic re-submitting with the same id after a timeout while the original call still runs; id generators restarting per process/thread; forgetting to retrieve or cancel a pending call before issuing the next.

Understand the failure class

Background: "Must be a positive integer", "Invalid value", "Unsupported": the invalid-argument-value error family, when a library rejects the value you pass — this error's family across 35 libraries.

Related errors

• call_id must be a nonzero uint64
• -32801
• timed out waiting for install lock {}
• A function call with ID {call_id} is already in progress
• trace call process id must be 16 bytes, got {}

AI-assisted analysis of BoundaryML/baml@bd85ce9dee (2026-09-12). Data as JSON: /api/errors/8b60f13233c7643d. Report an issue: GitHub.