为了更快的定位您的问题，请提供以下信息，谢谢
【详述】SELECT t0.exp_id AS exp_id, t0.grp_id AS grp_id, COVAR_SAMP(COALESCE(t0.zi, 0), COALESCE(t0.zi, 0)) AS abt_100011576_var_zi, COVAR_SAMP(COALESCE(t0.mu, 0), COALESCE(t0.mu, 0)) AS abt_100011576_var_mu, COVAR_SAMP(COALESCE(t0.zi, 0), COALESCE(t0.mu, 0)) AS abt_100011576_cov_zi_mu, COVAR_SAMP(COALESCE(t0.zi, 0), COALESCE(t1.uv, 0)) AS abt_100011576_cov_zi_uv, COVAR_SAMP(COALESCE(t0.mu, 0), COALESCE(t1.uv, 0)) AS abt_100011576_cov_mu_uv, COVAR_SAMP(COALESCE(t1.uv, 0), COALESCE(t1.uv, 0)) AS abt_100011576_var_uv, SUM(t0.zi) AS abt_100011576_sum_zi, SUM(t0.mu) AS abt_100011576_sum_mu, SUM(t1.uv) AS abt_100011576_sum_uv, SUM(t0.zi) / SUM(t0.mu) AS abt_100011576, COUNT(1) AS abt_100011576_bucket_num FROM(SELECT exp_id, grp_id, bucket_id, SUM(search_ymal_gmv) AS zi, SUM(search_eps_goods_num) AS mu FROM dm.dm_abt_exp_grp_bucket_search_top_source_trigger_1d WHERE exp_id = 100791211 AND grp_id IN (102059666, 102059661) AND dt BETWEEN ‘20260122’ AND ‘20260127’ GROUP BY exp_id, grp_id, bucket_id) AS t0 INNER JOIN (SELECT exp_id, grp_id, bucket_id, COUNT(DISTINCT user_sl_id) AS uv FROM dm.dm_abt_user_siteid_lang_detail_1d WHERE exp_id = 100791211 AND grp_id IN (102059666, 102059661) AND dt BETWEEN ‘20260122’ AND ‘20260127’ GROUP BY exp_id, grp_id, bucket_id) AS t1 ON t1.exp_id = t0.exp_id AND t1.grp_id = t0.grp_id AND t1.bucket_id = t0.bucket_id GROUP BY t0.exp_id, t0.grp_id;
上面的sql执行时间为6~7s，预计毫秒级
【背景】做过哪些操作？
TRACE TIMES OPTIMIZER后发现，主要耗时在CBO期间
【业务影响】
【是否存算分离】是
【StarRocks版本】3.5.10
【集群规模】例如：3fe（1 follower+2observer）+30be（fe与be混部）
【机器信息】fe 64C512G，be 32C256G
【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】
- fe.log/beINFO/相应截图
- 慢查询：
- Profile信息
- 并行度：show variables like ‘%parallel_fragment_exec_instance_num%’;
- pipeline是否开启：show variables like ‘%pipeline%’;
- be节点cpu和内存使用率截图
- 查询报错：
- query_dump，怎么获取query_dump文件
- be crash
- be.out
- 外表查询报错
- be.out和fe.warn.log