-
Notifications
You must be signed in to change notification settings - Fork 10
/
tpcds-benchmark-with-emrspark-ess.yaml
115 lines (115 loc) · 4.13 KB
/
tpcds-benchmark-with-emrspark-ess.yaml
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
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
apiVersion: "sparkoperator.k8s.io/v1beta2"
kind: SparkApplication
metadata:
name: tpcds-benchmark-emrspark-ess-10t
namespace: default
spec:
type: Scala
mode: cluster
image: registry.cn-beijing.aliyuncs.com/zf-spark/spark-2.4.5:for-tpc-ds-2
imagePullPolicy: Always
mainClass: com.databricks.spark.sql.perf.tpcds.TPCDS_Standalone
mainApplicationFile: "oss://<YOUR-BUCKET>/jars/spark-sql-perf-assembly-0.5.0-SNAPSHOT.jar"
arguments:
- "--dataset_location"
- "oss://<YOUR-BUCKET>/datasets/"
- "--output_location"
- "oss://<YOUR-BUCKET>/outputs/ack-pr-10t-emr-with-ess"
- "--iterations"
- "1"
- "--shuffle_partitions"
- "1000"
- "--scale_factor"
- "10000"
- "--regenerate_dataset"
- "false"
- "--regenerate_metadata"
- "false"
- "--only_generate_data_and_meta"
- "false"
- "--format"
- "parquet"
- "--query_exclude_list"
- "q14a,q14b,q67"
sparkVersion: 2.4.5
restartPolicy:
type: Never
hadoopConf:
"fs.oss.impl": "org.apache.hadoop.fs.aliyun.oss.AliyunOSSFileSystem"
"fs.oss.endpoint": "<YOUR-OSS-ENDPOINT>"
"fs.oss.accessKeyId": "<YOUR-ACCESS-KEY-ID>"
"fs.oss.accessKeySecret": "<YOUR-ACCESS-KEY-SECRET>"
hive.metastore.uris: thrift://service-hive-metastore.default:9083
hive.metastore.client.socket.timeout: 600s
sparkConf:
spark.eventLog.enabled: "true"
spark.eventLog.dir: "oss://<YOUR-BUCKET>/spark/eventlogs"
spark.driver.extraJavaOptions: "-XX:-PrintGC -XX:+PrintGCDetails -XX:+PrintGCTimeStamps -XX:+PrintGCDateStamps"
spark.driver.maxResultSize: 40g
spark.executor.extraJavaOptions: "-XX:MaxDirectMemorySize=6g -XX:-PrintGC -XX:+PrintGCDetails -XX:+PrintGCTimeStamps -XX:+PrintGCDateStamps"
spark.locality.wait.node: "0"
spark.locality.wait.process: "0"
spark.locality.wait.rack: "0"
spark.locality.wait: "0"
spark.memory.fraction: "0.8"
spark.memory.offHeap.enabled: "false"
spark.memory.offHeap.size: "17179869184"
spark.sql.adaptive.bloomFilterJoin.enabled: "false"
spark.sql.adaptive.enabled: "false"
spark.sql.analyze.column.async.delay: "200"
spark.sql.auto.reused.cte.enabled: "true"
spark.sql.broadcastTimeout: "3600"
spark.sql.columnVector.offheap.enabled: "false"
spark.sql.crossJoin.enabled: "true"
spark.sql.delete.optimizeInSubquery: "true"
spark.sql.dynamic.runtime.filter.bbf.enabled: "false"
spark.sql.dynamic.runtime.filter.enabled: "true"
spark.sql.dynamic.runtime.filter.exact.enabled: "true"
spark.sql.dynamic.runtime.filter.table.size.lower.limit: "1069547520"
spark.sql.dynamic.runtime.filter.table.size.upper.limit: "5368709120"
spark.sql.files.openCostInBytes: "34108864"
spark.sql.inMemoryColumnarStorage.compressed: "true"
spark.sql.join.preferNativeJoin: "false"
spark.sql.native.codecache: "true"
spark.sql.native.codegen.wholeStage: "false"
spark.sql.native.nativewrite: "false"
spark.sql.pkfk.optimize.enable: "true"
spark.sql.pkfk.riJoinElimination: "true"
spark.sql.shuffle.partitions: "1000"
spark.sql.simplifyDecimal.enabled: "true"
spark.sql.sources.parallelPartitionDiscovery.parallelism: "432"
spark.sql.sources.parallelPartitionDiscovery.threshold: "32"
spark.shuffle.reduceLocality.enabled: "false"
spark.shuffle.service.enabled: "true"
spark.dynamicAllocation.enabled: "false"
spark.shuffle.manager: org.apache.spark.shuffle.ess.EssShuffleManager
spark.ess.master.host: emr-rss-master.spark-rss
spark.ess.master.port: "9099"
spark.ess.push.data.buffer.size: 64k
spark.ess.push.data.max.inflight: "2048"
spark.ess.rpc.io.clientThreads: "8"
spark.ess.rpc.io.serverThreads: "8"
spark.ess.data.io.clientThreads: "8"
spark.ess.data.io.numConnectionsPerPeer: "8"
driver:
cores: 15
coreLimit: 15000m
memory: 50g
labels:
version: 2.4.5
serviceAccount: spark
env:
- name: TZ
value: "Asia/Shanghai"
executor:
cores: 4
coreLimit: 6000m
instances: 20
memory: 24g
memoryOverhead: 10g
deleteOnTermination: false
labels:
version: 2.4.5
env:
- name: TZ
value: "Asia/Shanghai"