What causes "Greenscale Logger Error - No logging URL found" in BerriAI/litellm?

Provide the Greenscale logging URL when configuring the callback (e.g. litellm.callbacks = ['greenscale'] plus the GreenscaleConfig/logging_url setup per docs)

BerriAI/litellm · error · Exception

Greenscale Logger Error - No logging URL found

Error message

Greenscale Logger Error - No logging URL found

What it means

Exception raised by the Greenscale custom logger when it is time to send the log event but self.greenscale_logging_url is None. The URL is expected to be injected (typically via the standard logging-callback config), and its absence means the callback cannot POST anywhere.

Source

Thrown at litellm/integrations/greenscale.py:48

            if type(end_time) is datetime and type(start_time) is datetime:
                data["invocationLatency"] = int((end_time - start_time).total_seconds() * 1000)

            # Add additional metadata keys to tags
            tags: Final = []
            metadata: Final = kwargs.get("litellm_params", {}).get("metadata", {})
            for key, value in metadata.items():
                if key.startswith("greenscale"):
                    if key == "greenscale_project":
                        data["project"] = value
                    elif key == "greenscale_application":
                        data["application"] = value
                    else:
                        tags.append({"key": key.replace("greenscale_", ""), "value": str(value)})

            data["tags"] = tags

            if self.greenscale_logging_url is None:
                raise Exception("Greenscale Logger Error - No logging URL found")

            response: Final = litellm.module_level_client.post(
                self.greenscale_logging_url,
                headers=self.headers,
                data=json.dumps(data, default=str),
            )
            if response.status_code != 200:
                print_verbose(f"Greenscale Logger Error - {response.text}, {response.status_code}")
            else:
                print_verbose(f"Greenscale Logger Succeeded - {response.text}")
        except Exception as e:
            print_verbose(f"Greenscale Logger Error - {e}, Stack trace: {traceback.format_exc()}")

View on GitHub (pinned to 6c2dcb801b)

Solutions

Provide the Greenscale logging URL when configuring the callback (e.g. litellm.callbacks = ['greenscale'] plus the GreenscaleConfig/logging_url setup per docs)
Enable LITELLM_LOG=DEBUG or verbose logging to confirm the handler state at init
Verify the callback object actually received the URL before first completion call

Example fix

# before
litellm.callbacks = ['greenscale']  # no url configured

# after
from litellm.integrations.greenscale import GreenscaleLogger
litellm.callbacks = [GreenscaleLogger(greenscale_logging_url='https://api.greenscale.ai/v1/logs')]

Defensive patterns

Strategy: validation

Validate before calling

logger = GreenscaleLogger(greenscale_logging_url=os.getenv("GREENSCALE_LOGGING_URL"))  # adjust to your litellm version's constructor
assert logger.greenscale_logging_url is not None, "GREENSCALE_LOGGING_URL must be set"

Prevention

Instantiate the logger with an explicit URL rather than relying on defaults
Enable verbose logging in staging to catch handler init problems
Assert required config at startup, not on first log event

When it happens

Trigger: Adding 'greenscale' to litellm.callbacks (or per-request greenscale params) without providing the Greenscale logging URL in the callback initialization/config. The log attempt then aborts before the HTTP call. Note the surrounding try/except prints via print_verbose, so in default verbose settings this may pass silently.

Common situations: Enabling the Greenscale integration by name only, assuming defaults exist; version changes in how callback init params are passed; typo'd env/config key so the URL never reaches the handler.

Related errors

AI-assisted analysis of BerriAI/litellm@6c2dcb801b (2026-08-15). Data as JSON: /api/errors/9993f25a46f633ba. Report an issue: GitHub.