为了更快的定位您的问题，请提供以下信息，谢谢
【详述】StarRocks查询paimon表array字段类型，导致be/cn Crash，必现。cn.out 有以下报错信息：
3.5.6 RELEASE (build 3ba6b8d distro centos arch x86_64)
query_id:423b081d-daeb-11f0-9883-f48e38c31d24, fragment_instance:423b081d-daeb-11f0-9883-f48e38c31d26
Hive file path: hdfs://xxx/paimon/warehouse/my_paimon_array_struct_test/bucket-0/data-ff422df1-2b89-4dc3-8ea9-84bced68276d-0.parquet, partition id: 0, length: 2143, offset: 0
*** Aborted at 1765936441 (unix time) try “date -d @1765936441” if you are using GNU date ***
PC: @ 0x8af86fa starrocks::parquet::LakeMetaHelper::_is_valid_type(starrocks::parquet::ParquetField const*, starrocks::TIcebergSchemaField const*, starrocks::TypeDescriptor const*) const
*** SIGSEGV (@0x30) received by PID 15957 (TID 0x7ff9e5910700) LWP(16528) from PID 48; stack trace: ***
@ 0x7ffb068e620b __pthread_once_slow
@ 0xbda9c54 google::(anonymous namespace)::FailureSignalHandler(int, siginfo_t*, void*)
@ 0x7ffb0632bd46 PosixSignals::chained_handler(int, siginfo*, void*) [clone .part.0]
@ 0x7ffb0632c79e JVM_handle_linux_signal
@ 0x7ff88e068460 (/tmp/libjfs-amd64.12.so+0x80745f)
start time: Wed Dec 17 09:54:55 CST 2025, server uptime: 09:54:55 up 685 days, 18:43, 2 users, load average: 262.90, 87.63, 41.38
Run with JEMALLOC_CONF: ‘percpu_arena:percpu,oversize_threshold:0,muzzy_decay_ms:5000,dirty_decay_ms:5000,metadata_thp:auto,background_thread:true,prof:true,prof_active:false’
WARNING: Logging before InitGoogleLogging() is written to STDERR
【复现步骤】
1.创建paimon表
CREATE TABLE my_paimon_array_struct_test
(
user_id BIGINT,
user_name STRING,
actions ARRAY<STRUCT<
timestamp: TIMESTAMP,
action_type: STRING,
page_url: STRING,
extra: MAP<STRING, STRING>>>,
last_active_date DATE
)
USING paimon
TBLPROPERTIES (
‘primary-key’ = ‘user_id’,
‘deletion-vectors.enabled’ = ‘true’,
‘file-index.bloom-filter.columns’ = ‘user_id’,
‘file-index.bloom-filter.user_id.items’ = ‘4000’,
‘bucket’ = ‘3’
) LOCATION ‘xxx/paimon/warehouse/my_paimon_array_struct_test’;
2.写入样例数据
INSERT INTO my_paimon_array_struct_test VALUES
(
10001,
‘张三’,
ARRAY(
NAMED_STRUCT(
‘timestamp’, CAST(‘2024-05-18 10:30:00’ AS TIMESTAMP),
‘action_type’, ‘login’,
‘page_url’, ‘/home’,
‘extra’, MAP(‘browser’, ‘Chrome’, ‘OS’, ‘Windows’)
),
NAMED_STRUCT(
‘timestamp’, CAST(‘2024-05-18 10:32:15’ AS TIMESTAMP),
‘action_type’, ‘view_product’,
‘page_url’, ‘/product/123’,
‘extra’, MAP(‘product_id’, ‘123’, ‘referer’, ‘/home’)
)
),
CAST(‘2024-05-18’ AS DATE)
);
3.使用StarRocks查询
select * from paimon_catalog.my_db.my_paimon_array_struct_test;
【是否存算分离】是
【StarRocks版本】3.5.6
【集群规模】3fe +2 be（fe与be混部）
【联系方式】