【 TiDB 使用环境】
生产环境
【 TiDB 版本】
6.5.0
【复现路径】
N/A
【遇到的问题：问题现象及影响】
https://docs.pingcap.com/zh/tidb/dev/tune-tiflash-performance#设置更大的执行并发度
执行计划有如下内容：
mysql> explain analyze select a, count(*) from t group by a;
±-------------------------------------±-------------±----------±-------------±--------------±-----------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------±-----------------------------------------------------------------------------------------------------------------------------±-------±-----+
| id | estRows | actRows | task | access object | execution info | operator info | memory | disk |
±-------------------------------------±-------------±----------±-------------±--------------±-----------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------±-----------------------------------------------------------------------------------------------------------------------------±-------±-----+
| TableReader_44 | 20.00 | 20 | root | | time:655.9ms, loops:2, cop_task: {num: 14, max: 0s, min: 0s, avg: 0s, p95: 0s, copr_cache_hit_ratio: 0.00} | data:ExchangeSender_43 | N/A | N/A |
| └─ExchangeSender_43 | 20.00 | 20 | mpp[tiflash] | | tiflash_task:{proc max:654.6ms, min:652.4ms, avg: 653.8ms, p80:654.6ms, p95:654.6ms, iters:11, tasks:3, threads:24} | ExchangeType: PassThrough | N/A | N/A |
| └─Projection_5 | 20.00 | 20 | mpp[tiflash] | | tiflash_task:{proc max:654.3ms, min:651.4ms, avg: 653.1ms, p80:654.3ms, p95:654.3ms, iters:11, tasks:3, threads:24} | test.t.a, Column#22 | N/A | N/A |
| └─Projection_39 | 20.00 | 20 | mpp[tiflash] | | tiflash_task:{proc max:654.3ms, min:651.4ms, avg: 653.1ms, p80:654.3ms, p95:654.3ms, iters:11, tasks:3, threads:24} | Column#22, test.t.a | N/A | N/A |
| └─HashAgg_40 | 20.00 | 20 | mpp[tiflash] | | tiflash_task:{proc max:654.3ms, min:651.4ms, avg: 653.1ms, p80:654.3ms, p95:654.3ms, iters:11, tasks:3, threads:24} | group by:test.t.a, funcs:sum(Column#29)->Column#22, funcs:firstrow(test.t.a)->test.t.a, stream_count: 8 | N/A | N/A |
| └─ExchangeReceiver_42 | 20.00 | 60 | mpp[tiflash] | | tiflash_task:{proc max:654.3ms, min:651.4ms, avg: 653.1ms, p80:654.3ms, p95:654.3ms, iters:11, tasks:3, threads:24} | stream_count: 8 | N/A | N/A |
| └─ExchangeSender_41 | 20.00 | 60 | mpp[tiflash] | | tiflash_task:{proc max:649ms, min:0s, avg: 216.3ms, p80:649ms, p95:649ms, iters:3, tasks:3, threads:24} | ExchangeType: HashPartition, Hash Cols: [name: test.t.a, collate: binary], stream_count: 8 | N/A | N/A |
| └─HashAgg_37 | 20.00 | 60 | mpp[tiflash] | | tiflash_task:{proc max:649ms, min:0s, avg: 216.3ms, p80:649ms, p95:649ms, iters:3, tasks:3, threads:24} | group by:test.t.a, funcs:count(1)->Column#29 | N/A | N/A |
| └─TableFullScan_26 | 600000000.00 | 600000000 | mpp[tiflash] | table:t | tiflash_task:{proc max:40ms, min:0s, avg: 13.3ms, p80:40ms, p95:40ms, iters:9386, tasks:3, threads:24}, tiflash_scan:{dtfile:{total_scanned_packs:73833, total_skipped_packs:174, total_scanned_rows:600000000, total_skipped_rows:1402537, total_rs_index_load_time: 0ms, total_read_time: 4ms}, total_create_snapshot_time: 0ms} | keep order:false | N/A | N/A |
±-------------------------------------±-------------±----------±-------------±--------------±-----------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------±-----------------------------------------------------------------------------------------------------------------------------±-------±-----+
9 rows in set (0.67 sec)
问题：
1）其中 tasks:3 代表这个 sql query 有 3 个 tiflash task？是根据什么拆分为 3 个？
2）threads:24，此时设置 tidb_max_tiflash_threads =24 ，那么这个代表是 mpp task 线程数最大用到 24 ，还是实际有其它的算法，上述文档只是凑巧 threads: 和 tidb_max_tiflash_threads 设置值相同？
【资源配置】
N/A
【附件：截图/日志/监控】
N/A