FATEBoard 多host时上传不同数据集但host数据集仍相同

[sealofyou]

Describe the bug
FATE1.11.3，多方训练
guest_save_path='/data/projects/fate/fate/upload/315/mnist_train_changeA' host_save_path=['/data/projects/fate/fate/upload/315/mnist_train_changeB', '/data/projects/fate/fate/upload/315/mnist_train_changeC'] guest=9999 host=[10000, 9000] arbiter=9999
其中10000, 9000在FATEBoard的数据集均显示为experiment.mnist_train_changeC

日志如下

# Reader
[INFO] [2025-11-28 10:20:47,814] [202511281020383644780] [445336:140139240544064] - [task_executor._run_] [line:158]: run reader_0 202511281020383644780_reader_0 0 on host 10000 task
[INFO] [2025-11-28 10:20:47,815] [202511281020383644780] [445336:140139240544064] - [task_executor._run_] [line:159]: component parameters on party:
{
    "dsl_version": 2,
    "initiator": {
        "role": "guest",
        "party_id": 9999
    },
    "role": {
        "guest": [
            9999
        ],
        "host": [
            10000,
            9000
        ]
    },
    "job_parameters": {
        "job_type": "train",
        "inheritance_info": {},
        "computing_engine": "STANDALONE",
        "federation_engine": "STANDALONE",
        "storage_engine": "STANDALONE",
        "engines_address": {
            "computing": {
                "cores_per_node": 4,
                "nodes": 4
            },
            "federation": {
                "cores_per_node": 4,
                "nodes": 4
            },
            "storage": {
                "cores_per_node": 4,
                "nodes": 4
            }
        },
        "federated_mode": "SINGLE",
        "task_parallelism": 1,
        "computing_partitions": 4,
        "federated_status_collect_type": "PUSH",
        "model_id": "arbiter-9999#guest-9999#host-9000_10000#model",
        "model_version": "202511281020383644780",
        "auto_retries": 0,
        "auto_retry_delay": 1,
        "eggroll_run": {
            "eggroll.session.processors.per.node": 1
        },
        "spark_run": {},
        "rabbitmq_run": {},
        "pulsar_run": {},
        "adaptation_parameters": {
            "task_nodes": 4,
            "task_cores_per_node": 1,
            "task_memory_per_node": 0,
            "request_task_cores": 1,
            "if_initiator_baseline": false
        },
        "task_conf": {},
        "roles": {
            "guest": [
                9999
            ],
            "host": [
                10000,
                9000
            ],
            "arbiter": [
                9999
            ]
        },
        "role_parameters": {}
    },
    "local": {
        "role": "host",
        "party_id": 10000
    },
    "module": "Reader",
    "CodePath": "Reader",
    "ComponentParam": {
        "table": {
            "name": "mnist_train_changeB",
            "namespace": "experiment"
        },
        "_name": "Reader#reader_0"
    },
    "ComponentParameterSource": "reader_0"
}

[INFO] [2025-11-28 10:20:47,996] [202511281020383644780] [445336:140139240544064] - [reader.save_table] [line:207]: source table name: mnist_train_changeB namespace: experiment engine: PATH
[INFO] [2025-11-28 10:20:47,996] [202511281020383644780] [445336:140139240544064] - [reader.save_table] [line:210]: destination table name: e8278626cc4311f087dc0242ac110005 namespace: output_data_202511281020383644780_reader_0_0 engine: PATH

# nn_0
[DEBUG] [2025-11-28 10:23:15,187] [202511281020383644780] [445525:140606553147200] - [data.load_dataset] [line:30]: use cached dataset, cached id /data/projects/fate/fate/upload/315/mnist_train_changeB

# eval_0未出现

请问如何解决，或者如何排查。目前训练是成功的。