Skip to content

vllm.v1.engine.exceptions

Classes:

EngineDeadError

Bases: VLLMServerError

Raised when the EngineCore dies. Unrecoverable.

Source code in vllm/v1/engine/exceptions.py 
14
15
16
17
18
19
20
21
22
23
class EngineDeadError(VLLMServerError):
    """Raised when the EngineCore dies. Unrecoverable."""

    def __init__(self, *args, suppress_context: bool = False, **kwargs):
        ENGINE_DEAD_MESSAGE = "EngineCore encountered an issue. See stack trace (above) for the root cause."  # noqa: E501

        super().__init__(ENGINE_DEAD_MESSAGE, *args, **kwargs)
        # Make stack trace clearer when using with LLMEngine by
        # silencing irrelevant ZMQError.
        self.__suppress_context__ = suppress_context

EngineGenerateError

Bases: VLLMServerError

Raised when a AsyncLLM.generate() fails. Recoverable.

Source code in vllm/v1/engine/exceptions.py 
 8
 9
10
11
class EngineGenerateError(VLLMServerError):
    """Raised when a AsyncLLM.generate() fails. Recoverable."""

    pass

EngineSleepingError

Bases: EngineUnhealthyError

Raised when the engine is intentionally sleeping or paused.

Source code in vllm/v1/engine/exceptions.py 
42
43
44
45
46
class EngineSleepingError(EngineUnhealthyError):
    """Raised when the engine is intentionally sleeping or paused."""

    def __init__(self, message: str = "", **details: Any):
        super().__init__(message, reason="sleeping", **details)

EngineUnhealthyError

Bases: VLLMServerError

Raised when the engine is alive but not ready to serve traffic.

Parameters:

  • message

    (str, default: '' ) –

    Human-readable description.

  • reason

    (str, default: 'unhealthy' ) –

    Machine-readable not-ready reason reported by /health_generate.

  • details

    (Any, default: {} ) –

    Extra JSON-serializable fields reported by /health_generate.

Source code in vllm/v1/engine/exceptions.py 
26
27
28
29
30
31
32
33
34
35
36
37
38
39
class EngineUnhealthyError(VLLMServerError):
    """Raised when the engine is alive but not ready to serve traffic.

    Args:
        message: Human-readable description.
        reason: Machine-readable not-ready reason reported by ``/health_generate``.
        details: Extra JSON-serializable fields reported by ``/health_generate``.

    """

    def __init__(self, message: str = "", reason: str = "unhealthy", **details: Any):
        super().__init__(message)
        self.reason = reason
        self.details = details