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
|
package spark.deploy
import collection.mutable.HashMap
import org.apache.hadoop.security.UserGroupInformation
import org.apache.hadoop.yarn.conf.YarnConfiguration
import org.apache.hadoop.conf.Configuration
import org.apache.hadoop.yarn.api.ApplicationConstants.Environment
import java.security.PrivilegedExceptionAction
/**
* Contains util methods to interact with Hadoop from spark.
*/
object SparkHadoopUtil {
val yarnConf = newConfiguration()
def getUserNameFromEnvironment(): String = {
// defaulting to env if -D is not present ...
val retval = System.getProperty(Environment.USER.name, System.getenv(Environment.USER.name))
// If nothing found, default to user we are running as
if (retval == null) System.getProperty("user.name") else retval
}
def runAsUser(func: (Product) => Unit, args: Product) {
runAsUser(func, args, getUserNameFromEnvironment())
}
def runAsUser(func: (Product) => Unit, args: Product, user: String) {
// println("running as user " + jobUserName)
UserGroupInformation.setConfiguration(yarnConf)
val appMasterUgi: UserGroupInformation = UserGroupInformation.createRemoteUser(user)
appMasterUgi.doAs(new PrivilegedExceptionAction[AnyRef] {
def run: AnyRef = {
func(args)
// no return value ...
null
}
})
}
// Note that all params which start with SPARK are propagated all the way through, so if in yarn mode, this MUST be set to true.
def isYarnMode(): Boolean = {
val yarnMode = System.getProperty("SPARK_YARN_MODE", System.getenv("SPARK_YARN_MODE"))
java.lang.Boolean.valueOf(yarnMode)
}
// Set an env variable indicating we are running in YARN mode.
// Note that anything with SPARK prefix gets propagated to all (remote) processes
def setYarnMode() {
System.setProperty("SPARK_YARN_MODE", "true")
}
def setYarnMode(env: HashMap[String, String]) {
env("SPARK_YARN_MODE") = "true"
}
// Return an appropriate (subclass) of Configuration. Creating config can initializes some hadoop subsystems
// Always create a new config, dont reuse yarnConf.
def newConfiguration(): Configuration = new YarnConfiguration(new Configuration())
}
|