I am attempting to build an on prem ml flow instance with scylla db as a backend data store. I can almost use this package as is, but I get the following error from mlflow:
Traceback (most recent call last):
2022-11-28 11:33:08 File "/opt/conda/lib/python3.9/site-packages/mlflow/cli.py", line 483, in server
2022-11-28 11:33:08 initialize_backend_stores(backend_store_uri, registry_store_uri, default_artifact_root)
2022-11-28 11:33:08 File "/opt/conda/lib/python3.9/site-packages/mlflow/server/handlers.py", line 268, in initialize_backend_stores
2022-11-28 11:33:08 _get_tracking_store(backend_store_uri, default_artifact_root)
2022-11-28 11:33:08 File "/opt/conda/lib/python3.9/site-packages/mlflow/server/handlers.py", line 247, in _get_tracking_store
2022-11-28 11:33:08 _tracking_store = _tracking_store_registry.get_store(store_uri, artifact_root)
2022-11-28 11:33:08 File "/opt/conda/lib/python3.9/site-packages/mlflow/tracking/_tracking_service/registry.py", line 39, in get_store
2022-11-28 11:33:08 return self._get_store_with_resolved_uri(resolved_store_uri, artifact_uri)
2022-11-28 11:33:08 File "/opt/conda/lib/python3.9/site-packages/mlflow/tracking/_tracking_service/registry.py", line 49, in _get_store_with_resolved_uri
2022-11-28 11:33:08 return builder(store_uri=resolved_store_uri, artifact_uri=artifact_uri)
2022-11-28 11:33:08 File "/opt/conda/lib/python3.9/site-packages/dynaflow/tracking_store.py", line 159, in __init__
2022-11-28 11:33:08 if not BaseEntry.exists():
2022-11-28 11:33:08 File "/opt/conda/lib/python3.9/site-packages/pynamodb/models.py", line 762, in exists
2022-11-28 11:33:08 cls._get_connection().describe_table()
2022-11-28 11:33:08 File "/opt/conda/lib/python3.9/site-packages/pynamodb/connection/table.py", line 294, in describe_table
2022-11-28 11:33:08 return self.connection.describe_table(self.table_name)
2022-11-28 11:33:08 File "/opt/conda/lib/python3.9/site-packages/pynamodb/connection/base.py", line 787, in describe_table
2022-11-28 11:33:08 raise TableError("Unable to describe table: {}".format(e), e)
2022-11-28 11:35:10 pynamodb.exceptions.TableError: Unable to describe table: Could not connect to the endpoint URL: "https://localhost:8000/"
I am attempting to build an on prem ml flow instance with scylla db as a backend data store. I can almost use this package as is, but I get the following error from mlflow: