doris1.0.0 Unique模型 + 开启向量化执行出现内存泄漏

[980150976]

表模型：unique

CREATE TABLE pack_test(start_timebigint(20) NULL COMMENT "开始采集时间",end_timebigint(20) NULL COMMENT "结束采集时间",_idvarchar(255) NULL COMMENT "_id",dtdate NULL COMMENT "日期，例如2022-05-10",artifact_id int(11) NULL COMMENT "artifact_id", xx ) ENGINE=OLAP UNIQUE KEY(start_time, end_time, _id, dt) COMMENT "OLAP" PARTITION BY RANGE(dt) (PARTITION p2019 VALUES [('0000-01-01'), ('2020-01-01')), PARTITION p2020 VALUES [('2020-01-01'), ('2021-01-01')), PARTITION p2021 VALUES [('2021-01-01'), ('2022-01-01')), PARTITION p2022 VALUES [('2022-01-01'), ('2023-01-01'))) DISTRIBUTED BY HASH(_id) BUCKETS 16 PROPERTIES ( "replication_allocation" = "tag.location.default: 3", "bloom_filter_columns" = "duid, puid, name, _id, artifact_id, url", "dynamic_partition.enable" = "true", "dynamic_partition.time_unit" = "month", "dynamic_partition.time_zone" = "Asia/Shanghai", "dynamic_partition.start" = "-2147483648", "dynamic_partition.end" = "3", "dynamic_partition.prefix" = "p", "dynamic_partition.replication_allocation" = "tag.location.default: 3", "dynamic_partition.buckets" = "16", "dynamic_partition.create_history_partition" = "false", "dynamic_partition.history_partition_num" = "-1", "dynamic_partition.hot_partition_num" = "0", "dynamic_partition.reserved_history_periods" = "NULL", "dynamic_partition.start_day_of_month" = "1", "in_memory" = "false", "storage_format" = "V2" )

表总数据量1000w左右，占空间11G左右

set global enable_vectorized_engine = true
set global batch_size = 4096;

集群内存>1T

不断查询后，内存持续飙升，直到BE挂掉

尝试用明细模型，没有这个问题！！！

[980150976]

doris版本：https://www.apache.org/dyn/closer.cgi/incubator/doris/1.0/1.0.0-incubating/apache-doris-1.0.0-incubating-bin.tar.gz， 未采用源码编译；不知道和操作系统环境是否有关系，还待确定。可以确定的是这个问题是必现。

[dataalive]

感谢反馈，这个问题我们正在压测和解决。

[980150976]

补充：有些字段存在大量的null

[kpfly]

补充：有些字段存在大量的null

Maybe you can try doris 1.1 ，some memory leak issues have been solved. Thanks.
https://github.com/apache/doris/releases/tag/1.1.0-rc03