[doris 3.0.2] k8s部署，集群外streamload上传失败

[Clara12062]

Hi～ k8s部署的doris集群， streanload的使用方式只限于集群内吗？

问题描述

当前doris集群使用存算一体部署到k8s集群中， 对于的svc如下：

NAME                        TYPE        CLUSTER-IP       EXTERNAL-IP   PORT(S)                                        AGE
doris-cluster-be-internal   ClusterIP   None             <none>        9050/TCP,8060/TCP                              45d
doris-cluster-be-service    ClusterIP   172.16.165.222   <none>        9060/TCP,8040/TCP,9050/TCP,8060/TCP            64d
doris-cluster-fe-internal   ClusterIP   None             <none>        9030/TCP,9020/TCP,9010/TCP,8030/TCP            60d
doris-cluster-fe-service    ClusterIP   172.16.235.178   <none>        8030/TCP,9020/TCP,9030/TCP,9010/TCP,9060/TCP   60d

在集群外配置了LB. 当前访问:9030/8030端口都可以正常访问， 查看backends信息如下：

MySQL [(none)]> SHOW BACKENDS;
+-----------+-----------------------------------------------------------------------+---------------+--------+----------+----------+--------------------+---------------------+---------------------+-------+----------------------+-----------+------------------+-------------------+---------------+---------------+---------+----------------+--------------------+--------------------------+--------+-----------------------------+-------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------+-------------------------+----------+----------+----------+
| BackendId | Host                                                                  | HeartbeatPort | BePort | HttpPort | BrpcPort | ArrowFlightSqlPort | LastStartTime       | LastHeartbeat       | Alive | SystemDecommissioned | TabletNum | DataUsedCapacity | TrashUsedCapacity | AvailCapacity | TotalCapacity | UsedPct | MaxDiskUsedPct | RemoteUsedCapacity | Tag                      | ErrMsg | Version                     | Status                                                                                                                                                                                                      | HeartbeatFailureCounter | NodeRole | CpuCores | Memory   |
+-----------+-----------------------------------------------------------------------+---------------+--------+----------+----------+--------------------+---------------------+---------------------+-------+----------------------+-----------+------------------+-------------------+---------------+---------------+---------+----------------+--------------------+--------------------------+--------+-----------------------------+-------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------+-------------------------+----------+----------+----------+
| 34102     | doris-cluster-be-0.doris-cluster-be-internal.becbvd.svc.cluster.local | 9050          | 9060   | 8040     | 8060     | -1                 | 2025-03-27 04:11:56 | 2025-05-09 09:52:34 | true  | false                | 1096      | 23.670 GB        | 0.000             | 475.911 GB    | 500.000 GB    | 4.82 %  | 4.82 %         | 0.000              | {"location" : "default"} |        | doris-3.0.2-rc03-c21b9f5bce | {"lastSuccessReportTabletsTime":"2025-05-09 09:51:49","lastStreamLoadTime":-1,"isQueryDisabled":false,"isLoadDisabled":false,"isActive":true,"currentFragmentNum":0,"lastFragmentUpdateTime":1746784354435} | 0                       | mix      | 6        | 16.00 GB |
| 172883    | doris-cluster-be-1.doris-cluster-be-internal.becbvd.svc.cluster.local | 9050          | 9060   | 8040     | 8060     | -1                 | 2025-03-27 02:11:57 | 2025-05-09 09:52:34 | true  | false                | 1065      | 20.881 GB        | 0.000             | 478.804 GB    | 500.000 GB    | 4.24 %  | 4.24 %         | 0.000              | {"location" : "default"} |        | doris-3.0.2-rc03-c21b9f5bce | {"lastSuccessReportTabletsTime":"2025-05-09 09:51:37","lastStreamLoadTime":-1,"isQueryDisabled":false,"isLoadDisabled":false,"isActive":true,"currentFragmentNum":0,"lastFragmentUpdateTime":1746784354190} | 0                       | mix      | 6        | 16.00 GB |
| 173136    | doris-cluster-be-2.doris-cluster-be-internal.becbvd.svc.cluster.local | 9050          | 9060   | 8040     | 8060     | -1                 | 2025-03-26 18:17:56 | 2025-05-09 09:52:34 | true  | false                | 1063      | 21.256 GB        | 0.000             | 478.354 GB    | 500.000 GB    | 4.33 %  | 4.33 %         | 0.000              | {"location" : "default"} |        | doris-3.0.2-rc03-c21b9f5bce | {"lastSuccessReportTabletsTime":"2025-05-09 09:51:59","lastStreamLoadTime":-1,"isQueryDisabled":false,"isLoadDisabled":false,"isActive":true,"currentFragmentNum":0,"lastFragmentUpdateTime":1746784354504} | 0                       | mix      | 6        | 16.00 GB |
+-----------+-----------------------------------------------------------------------+---------------+--------+----------+----------+--------------------+---------------------+---------------------+-------+----------------------+-----------+------------------+-------------------+---------------+---------------+---------+----------------+--------------------+--------------------------+--------+-----------------------------+-------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------+-------------------------+----------+----------+----------+

在集群外， 访问lb-ip:8030使用streamload导入数据报错：

# curl --location-trusted -u admin:password -T test.csv   -H "Expect:100-continue"   http://lb-ip:8030/api/flow_detect_sandbox/flow_cross_province_tbl/_stream_load
curl: (6) Could not resolve host: doris-cluster-be-0.doris-cluster-be-internal.becbvd.svc.cluster.local; Unknown error

我理解的fe的作用是代理， 然后将请求重定向到be的话， 会是interval的pod-ip， 这个domain：doris-cluster-be-0.doris-cluster-be-internal.becbvd.svc.cluster.local集群外不可访问， 所以这种情况下： doris部署到k8s集群内， 集群外走stream load通过fe-ip:8030无法导入数据对吗？

或者直接给doris-be也添加LB， 直接访问BE的pod是么？