存算一体 3.5.1 版本,采用倒排索引插入数据后be宕机

为了更快的定位您的问题,请提供以下信息,谢谢
【详述】存算一体 3.5.1 版本,采用倒排索引的表,插入数据后be宕机
【背景】

1. 建表语句:

CREATE TABLE session_table_2025_03 (
session_id bigint NOT NULL,
session_source string,
game_user_id string,
send_info string,
service_name string,
chat_message string,
send_time datetime,
session_type string,
message_type string,
INDEX idx (chat_message) USING GIN(“parser” = “chinese”)
) ENGINE=OLAP
DUPLICATE KEY(session_id)
DISTRIBUTED BY HASH(session_id) BUCKETS 1
PROPERTIES (
“replicated_storage” = “false”,
“replication_num” = “1”
);

2. 设置fe参数

ADMIN SET FRONTEND CONFIG (“enable_experimental_gin” = “true”);

3. stream load 或者 insert 导入数据,数据导入后,过一会儿就be宕机

4. be.out日志

3.5.1 RELEASE (build d550b36)
query_id:00000000-0000-0000-0000-000000000000, fragment_instance:00000000-0000-0000-0000-000000000000
*** Aborted at 1760317080 (unix time) try “date -d @1760317080” if you are using GNU date ***
PC: @ 0xb277696 lucene::analysis::cjk::CJKTokenizer::next(lucene::analysis::Token*)
*** SIGSEGV (@0x14ad00000454) received by PID 1897779 (TID 0x14ad263f8640) LWP(1898608) from PID 1108; stack trace: ***
@ 0x14adf370d19f __pthread_once_slow
@ 0xb708e94 google::(anonymous namespace)::FailureSignalHandler(int, siginfo_t*, void*)
@ 0x14adf37101d0 (/usr/lib64/libpthread-2.32.so+0x141cf)
@ 0xb277696 lucene::analysis::cjk::CJKTokenizer::next(lucene::analysis::Token*)
@ 0xb2285e1 lucene::index::DocumentsWriter::ThreadState::FieldData::invertField(lucene::document::Field*, lucene::analysis::Analyzer*, int)
@ 0xb228bd3 lucene::index::DocumentsWriter::ThreadState::FieldData::processField(lucene::analysis::Analyzer*)
@ 0xb228f4e lucene::index::DocumentsWriter::ThreadState::processDocument(lucene::analysis::Analyzer*)
@ 0xb21f62c lucene::index::DocumentsWriter::addDocument(lucene::document::Document*, lucene::analysis::Analyzer*)
@ 0xb1ea4a7 lucene::index::IndexWriter::addDocument(lucene::document::Document*, lucene::analysis::Analyzer*)
@ 0x7d414cf starrocks::CLuceneInvertedWriterImpl<(starrocks::LogicalType)17>::add_values(void const*, unsigned long)
@ 0x7b09d7b starrocks::ScalarColumnWriter::append(unsigned char const*, unsigned char const*, unsigned long, bool)
@ 0x7b0c1c4 starrocks::ScalarColumnWriter::append(starrocks::Column const&)
@ 0x7b0c34e starrocks::StringColumnWriter::append(starrocks::Column const&)
@ 0x72a516d starrocks::SegmentWriter::append_chunk(starrocks::Chunk const&)
@ 0x7db6c1a starrocks::VerticalRowsetWriter::add_columns(starrocks::Chunk const&, std::vector<unsigned int, std::allocator > const&, bool, std::vector<unsigned long, std::allocator > const&)
@ 0x7db721a starrocks::VerticalRowsetWriter::add_columns(starrocks::Chunk const&, std::vector<unsigned int, std::allocator > const&, bool)
@ 0x7d88fbe starrocks::VerticalCompactionTask::_compact_data(bool, int, std::vector<unsigned int, std::allocator > const&, starrocks::Schema const&, starrocks::TabletReader*, starrocks::RowsetWriter*, starrocks::RowSourceMaskBuffer*, std::vector<starrock@
@ 0x7d89aae starrocks::VerticalCompactionTask::_compact_column_group(bool, int, std::vector<unsigned int, std::allocator > const&, starrocks::RowsetWriter*, starrocks::RowSourceMaskBuffer*, std::vector<starrocks::RowSourceMask, std::allocator<starrocks::@
@ 0x7d8a2b2 starrocks::VerticalCompactionTask::_vertical_compaction_data(starrocks::Statistics*)
@ 0x7d8ab3e starrocks::VerticalCompactionTask::run_impl()
@ 0x7d833ee starrocks::CompactionTask::run()
@ 0x74aa58f std::_Function_handler<void (), starrocks::CompactionManager::submit_compaction_task(starrocks::CompactionCandidate const&)::{lambda()#1}>::_M_invoke(std::_Any_data const&)
@ 0x40b7728 starrocks::ThreadPool::dispatch_thread()
@ 0x40ae9b0 starrocks::thread::supervise_thread(void*)
@ 0x14adf37053fb start_thread
@ 0x14adf1d1be83 __GI___clone
start time: Mon Oct 13 09:00:39 AM CST 2025, server uptime: 09:00:39 up 100 days, 19:01, 2 users, load average: 0.15, 0.18, 0.23
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’
Ignored unknown config: datacache_disk_path
Ignored unknown config: disable_column_pool

【业务影响】无法使用倒排索引
【是否存算分离】存算一体
【StarRocks版本】3.5.1
【集群规模】1fe + 1be
【机器信息】
【联系方式】
【附件】