kudu-user mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From "fengbaoli@uce.cn" <fengba...@uce.cn>
Subject question about kudu performance!
Date Wed, 04 Apr 2018 02:38:00 GMT
The component version is:
   - CDH 5.14
   - Kudu 1.6(2)FrameworkThe size of the kudu cluster(total 10 machines,256G mem,16*1.2T sas
disk):-3 master node-7 tablet server nodeIndependent deployment of master and tablet server,but
Yarn nodemanger and tablet server are deployed together (3) kudu parameter :maintenance_manager_num_threads=24(The
number of data directories for each machine disk is 8)memory_limit_hard_bytes=150GI have a
performance problem: every 2-3 weeks, clusters start to make MajorDeltaCompactionOp, when
kudu performs insert and update performance decreases, when the data is written, the update
operation almost stops.Is it possible to adjust the --memory_limit_hard_bytes parameter to
256G*80%(my yarn nm and TS are deployed together)?Can we adjust the parameter --tablet_history_max_age_sec
to shorten the MajorDeltaCompactionOp interval?Can you give me some suggestions to optimize
this performance problem?thanks!

大数据中心     冯宝利
View raw message