数据湖iceberg的统计信息写入失败, 统计信息表的主键超长了

💬 StarRocks 用户问答
#1

为了更快的定位您的问题，请提供以下信息，谢谢
【详述】问题详细描述
统计信息采集报错

mysql> ANALYZE FULL TABLE lake.saas_datalake.dwd_adv_release_effect_info_f;
+--------------------------------------------------+---------+----------+---------------------------------------------------+
| Table                                            | Op      | Msg_type | Msg_text                                          |
+--------------------------------------------------+---------+----------+---------------------------------------------------+
| lake.saas_datalake.dwd_adv_release_effect_info_f | analyze | error    | sz-pre-node00: primary key size exceed the limit. |
+--------------------------------------------------+---------+----------+---------------------------------------------------+

统计信息表的建表语句

 CREATE TABLE `external_column_statistics` (
  `table_uuid` varchar(65530) NOT NULL COMMENT "",
  `partition_name` varchar(65530) NOT NULL COMMENT "",
  `column_name` varchar(65530) NOT NULL COMMENT "",
  `catalog_name` varchar(65530) NOT NULL COMMENT "",
  `db_name` varchar(65530) NOT NULL COMMENT "",
  `table_name` varchar(65530) NOT NULL COMMENT "",
  `row_count` bigint(20) NOT NULL COMMENT "",
  `data_size` bigint(20) NOT NULL COMMENT "",
  `ndv` hll NOT NULL COMMENT "",
  `null_count` bigint(20) NOT NULL COMMENT "",
  `max` varchar(1048576) NOT NULL COMMENT "",
  `min` varchar(1048576) NOT NULL COMMENT "",
  `update_time` datetime NOT NULL COMMENT ""
) ENGINE=OLAP 
PRIMARY KEY(`table_uuid`, `partition_name`, `column_name`)
DISTRIBUTED BY HASH(`table_uuid`, `partition_name`, `column_name`) BUCKETS 10 
PROPERTIES (
"compression" = "LZ4",
"enable_persistent_index" = "true",
"fast_schema_evolution" = "true",
"replicated_storage" = "true",
"replication_num" = "1"
);

湖表分区

PARTITION BY (yuce_tenant_id, data_source)

统计信息示例数据(把统计信息表的主键去掉后)
分区字段比较长 yuce_tenant_id=yuce_512405/data_source=拼多多RPA

+---------------------------------------------------------------------------------------+---------------------+-----------------------------------------------------+--------------+---------------+-------------------------------+-----------+-----------+------+------------+------------------------------+------------------------------+---------------------+
| table_uuid                                                                            | column_name         | partition_name                                      | catalog_name | db_name       | table_name                    | row_count | data_size | ndv  | null_count | max                          | min                          | update_time         |
+---------------------------------------------------------------------------------------+---------------------+-----------------------------------------------------+--------------+---------------+-------------------------------+-----------+-----------+------+------------+------------------------------+------------------------------+---------------------+
| lake.saas_datalake.dwd_adv_release_effect_info_f.5603cd1b-3795-4aa8-9a50-a8fbd2f1222a | account_name        | yuce_tenant_id=yuce_512405/data_source=拼多多RPA    | lake         | saas_datalake | dwd_adv_release_effect_info_f |        31 |         0 | NULL |         31 |                              |                              | 2026-01-21 17:33:48 |
| lake.saas_datalake.dwd_adv_release_effect_info_f.5603cd1b-3795-4aa8-9a50-a8fbd2f1222a | click               | yuce_tenant_id=yuce_512405/data_source=拼多多RPA    | lake         | saas_datalake | dwd_adv_release_effect_info_f |        31 |       248 | NULL |          0 | 28007                        | 8238                         | 2026-01-21 17:33:48 |

能不能把partitionName改成使用partitionId
否则统计信息写不进来了

【背景】做过哪些操作？
【业务影响】
【是否存算分离】
【StarRocks版本】例如：3.3.17
【集群规模】例如：3fe（1 follower+2observer）+5be（fe与be混部）
【机器信息】CPU虚拟核/内存/网卡，例如：48C/64G/万兆
【联系方式】为了在解决问题过程中能及时联系到您获取一些日志信息，请补充下您的联系方式，例如：社区群4-小李或者邮箱，谢谢
【附件】

京ICP备2022026421号-1