Doris 3.0.5-rc01 stream load doris_be cpu high usage than Drois 2.1.6-rc04 #51010
Unanswered
zicat
asked this question in
A - General / Q&A
Replies: 3 comments 3 replies
-
将表的参数"group_commit_interval_ms" = "20000"调整成"group_commit_interval_ms" = "1000" |
Beta Was this translation helpful? Give feedback.
0 replies
-
|
Beta Was this translation helpful? Give feedback.
2 replies
-
The issue #51103 |
Beta Was this translation helpful? Give feedback.
1 reply
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Uh oh!
There was an error while loading. Please reload this page.
Uh oh!
There was an error while loading. Please reload this page.
-
你好维护者:




近期我在从事将公司的doris版本从2.1.6-rc04升级到Doris 3.0.5-rc01相关工作(3.0.5-bin-x64-noavx2),部署模式为存算一体模式。
在升级的过程发现doris be进程的CPU使用率远高于旧版本。为了验证这一点我用5台server搭建了一个2.1.6-rc04集群,每台server上部署了一个doris be和doris fe进程,并创建了如下表进行stream load的导入
CREATE TABLE
rtc_freeze_1mislice(
align_tsbigint NULL,
namevarchar(255) NULL,
productvarchar(255) NULL,
vidbigint NULL,
devicevarchar(255) NULL,
regionvarchar(255) NULL,
countryvarchar(255) NULL,
netvarchar(255) NULL,
idcvarchar(255) NULL,
freeze_timebigint NULL,
user_cntbigint NULL,
domainvarchar(255) NULL,
datedate NULL,
hourint NULL ) ENGINE=OLAP DUPLICATE KEY(
align_ts,
name,
product) PARTITION BY RANGE(
date)() DISTRIBUTED BY RANDOM BUCKETS AUTO PROPERTIES ( "replication_allocation" = "tag.location.default: 5", "min_load_replica_num" = "-1", "is_being_synced" = "false", "dynamic_partition.enable" = "true", "dynamic_partition.time_unit" = "DAY", "dynamic_partition.time_zone" = "UTC", "dynamic_partition.start" = "-29", "dynamic_partition.end" = "2", "dynamic_partition.prefix" = "p", "dynamic_partition.replication_allocation" = "tag.location.default: 5", "dynamic_partition.buckets" = "10", "dynamic_partition.create_history_partition" = "true", "dynamic_partition.history_partition_num" = "-1", "dynamic_partition.hot_partition_num" = "0", "dynamic_partition.reserved_history_periods" = "NULL", "dynamic_partition.storage_policy" = "", "storage_medium" = "hdd", "storage_format" = "V2", "inverted_index_storage_format" = "V1", "compression" = "ZSTD", "light_schema_change" = "true", "disable_auto_compaction" = "false", "enable_single_replica_compaction" = "true", "group_commit_interval_ms" = "20000", "group_commit_data_bytes" = "655360000" );
CPU和stream load相关指标如下:
然后,我将该集群升级至3.0.5-rc01版本,用相同的数据进行stream load导入,
CPU和stream load相关指标如下:
可以看到,CPU使用率有明显的提升,想咨询下 3.0.5和2.1.6版本在数据导入上有什么改动会影响CPU吗,我查看了相关的release note没有找到相关的线索
Beta Was this translation helpful? Give feedback.
All reactions