티스토리 뷰
Yarn의 설정은 다음과 같다.
<YARN의 메모리 설정>
yarn.nodemanager.resource.memory-mb
- 노드의 메모리 크기 설정
yarn.nodemanager.resource.cpu-vcores
- 노드의 코어 개수 설정
<컨테이너의 메모리 설정>
yarn.scheduler.minimum-allocation-mb
yarn.scheduler.maximum-allocation-mb
- RM이 컨테이너당 할당하는 최소, 최대 메모리
yarn.scheduler.minimum-allocation-vcores
yarn.scheduler.maximum-allocation-vcores
- RM이 컨테이너당 할당하는 최소, 최대 코어
<MR의 메모리 설정>
매퍼 설정은 다음과 같다
a. mapreduce.map.memory.mb
b. mapreduce.map.cpu.vcores
c. mapreduce.map.java.opts
리듀서 설정은 다음과 같다.
d. mapreduce.reduce.memory.mb
e. mapreduce.reduce.cpu.vcores
f. mapreduce.reduce.java.opts
- vcores 설정은 1이 기본. 맵, 리듀서에서 멀티스레딩을 이용할 경우가 적음
- 리듀서의 메모리 설정은 맵의 2~4배로 설정하는 것이 일반적이다.
- java.opts 는 메모리 설정의 0.8배로 설정하는 것이 일반적이다.
g. yarn.app.mapreduce.am.resource.mb
- 맵리듀서를 위한 애플리케이션 마스터의 메모리 설정
'빅데이터 > hadoop' 카테고리의 다른 글
[hdfs] HDFS 디렉토리 최대값 설정 변경 (0) | 2018.12.10 |
---|---|
[hadoop] 세컨더리네임노드(secondary namenode) (0) | 2018.09.14 |
[hadoop] 맵리듀스 동작 순서 (0) | 2018.04.17 |
[hadoop] Ambari, Avro (0) | 2018.03.29 |
[hadoop] shuffle 단계 메모리 설정 정보 (0) | 2018.03.26 |
- Total
- Today
- Yesterday
- 오류
- SQL
- 알고리즘
- java
- HDFS
- Python
- HIVE
- 하둡
- 파이썬
- 하이브
- error
- ubuntu
- 다이나믹
- build
- nodejs
- airflow
- mysql
- Tez
- emr
- SPARK
- yarn
- S3
- Hadoop
- hbase
- oozie
- 백준
- bash
- Linux
- 정올
- AWS
일 | 월 | 화 | 수 | 목 | 금 | 토 |
---|---|---|---|---|---|---|
1 | 2 | 3 | 4 | 5 | ||
6 | 7 | 8 | 9 | 10 | 11 | 12 |
13 | 14 | 15 | 16 | 17 | 18 | 19 |
20 | 21 | 22 | 23 | 24 | 25 | 26 |
27 | 28 | 29 | 30 |