You are viewing a plain text version of this content. The canonical link for it is here.
Posted to notifications@iotdb.apache.org by "Liuxuxin (Jira)" <ji...@apache.org> on 2022/03/10 02:33:00 UTC

[jira] [Commented] (IOTDB-2636) [compaction] Compaction affects query performance

    [ https://issues.apache.org/jira/browse/IOTDB-2636?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=17503957#comment-17503957 ] 

Liuxuxin commented on IOTDB-2636:
---------------------------------

在 IoTDB-Benchmark 中开启 debug 查询,比例为 1/1000,得到查询时的debug日志。对日志分析得到:benchmark查询的文件均为0层文件,没有 1 层文件。因此开启合并后查询性能下降可能是合并占用 CPU 和 IO 所致。

> [compaction] Compaction affects query performance
> -------------------------------------------------
>
>                 Key: IOTDB-2636
>                 URL: https://issues.apache.org/jira/browse/IOTDB-2636
>             Project: Apache IoTDB
>          Issue Type: Bug
>          Components: Core/Compaction
>    Affects Versions: 0.13.0-SNAPSHOT
>            Reporter: 刘珍
>            Assignee: Liuxuxin
>            Priority: Critical
>         Attachments: image-2022-03-01-14-30-59-082.png, image-2022-03-01-14-37-30-548.png, image-2022-03-01-14-44-37-559.png, load-file.log
>
>
> master-pre3
> 测试环境:
> 私有云8C32G,16Gheap,无乱序,长测配置。
> 1. master 关合并/开合并对比
> cwt代表参数 merge_write_throughput_mb_per_sec
> master开合并会导致查询性能下降约20%。
>  !image-2022-03-01-14-30-59-082.png! 
> 2. master-pre3和0.12.5-pre6开合并对比
> merge_write_throughput_mb_per_sec均为8
> master总体比0.12慢70%
>  !image-2022-03-01-14-37-30-548.png! 
> 这个查询对应的写入性能(作为查询基于的数据量的参照),master查询的数据不大于0.12.5:
>  !image-2022-03-01-14-44-37-559.png! 



--
This message was sent by Atlassian Jira
(v8.20.1#820001)