Apache/hadoop
1
0
Fork 0
Code Issues Pull Requests Packages Projects Releases Wiki Activity

HADOOP-12317. Applications fail on NM restart on some linux distro because NM container recovery declares AM container as LOST (adhoot via rkanter)

Browse Source 
(cherry picked from commit 1e06299df82b98795124fe8a33578c111e744ff4)
This commit is contained in:
Robert Kanter 2015-08-19 19:00:51 -07:00
parent 62dfe1b7e4
commit acf241242d
3 changed files with 51 additions and 3 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

4
hadoop-common-project/hadoop-common/CHANGES.txt
View File

@ -566,6 +566,10 @@ Release 2.8.0 - UNRELEASED
HADOOP-12322. typos in rpcmetrics.java. (Anu Engineer via
Arpit Agarwal)
HADOOP-12317. Applications fail on NM restart on some linux distro
because NM container recovery declares AM container as LOST
(adhoot via rkanter)
Release 2.7.2 - UNRELEASED
INCOMPATIBLE CHANGES

11
hadoop-common-project/hadoop-common/src/main/java/org/apache/hadoop/util/Shell.java
View File

@ -212,13 +212,18 @@ public static String[] getReadlinkCommand(String link) {
public static String[] getCheckProcessIsAliveCommand(String pid) {
return Shell.WINDOWS ?
new String[] { Shell.WINUTILS, "task", "isAlive", pid } :
new String[] { "kill", "-0", isSetsidAvailable ? "-" + pid : pid };
isSetsidAvailable ?
new String[] { "kill", "-0", "--", "-" + pid } :
new String[] { "kill", "-0", pid };
}
/** Return a command to send a signal to a given pid */
public static String[] getSignalKillCommand(int code, String pid) {
return Shell.WINDOWS ? new String[] { Shell.WINUTILS, "task", "kill", pid } :
new String[] { "kill", "-" + code, isSetsidAvailable ? "-" + pid : pid };
return Shell.WINDOWS ?
new String[] { Shell.WINUTILS, "task", "kill", pid } :
isSetsidAvailable ?
new String[] { "kill", "-" + code, "--", "-" + pid } :
new String[] { "kill", "-" + code, pid };
}
public static final String ENV_NAME_REGEX = "[A-Za-z_][A-Za-z0-9_]*";

39
hadoop-common-project/hadoop-common/src/test/java/org/apache/hadoop/util/TestShell.java
View File

@ -18,6 +18,7 @@
package org.apache.hadoop.util;
import junit.framework.TestCase;
import org.junit.Assert;
import java.io.BufferedReader;
import java.io.File;
@ -150,6 +151,44 @@ public void testShellCommandTimerLeak() throws Exception {
System.err.println("after: " + timersAfter);
assertEquals(timersBefore, timersAfter);
}
public void testGetCheckProcessIsAliveCommand() throws Exception {
String anyPid = "9999";
String[] checkProcessAliveCommand = Shell.getCheckProcessIsAliveCommand(
anyPid);
String[] expectedCommand;
if (Shell.WINDOWS) {
expectedCommand =
new String[]{ Shell.WINUTILS, "task", "isAlive", anyPid };
} else if (Shell.isSetsidAvailable) {
expectedCommand = new String[]{ "kill", "-0", "--", "-" + anyPid };
} else {
expectedCommand = new String[]{"kill", "-0", anyPid};
}
Assert.assertArrayEquals(expectedCommand, checkProcessAliveCommand);
}
public void testGetSignalKillCommand() throws Exception {
String anyPid = "9999";
int anySignal = 9;
String[] checkProcessAliveCommand = Shell.getSignalKillCommand(anySignal,
anyPid);
String[] expectedCommand;
if (Shell.WINDOWS) {
expectedCommand =
new String[]{ Shell.WINUTILS, "task", "kill", anyPid };
} else if (Shell.isSetsidAvailable) {
expectedCommand =
new String[]{ "kill", "-" + anySignal, "--", "-" + anyPid };
} else {
expectedCommand =
new String[]{ "kill", "-" + anySignal, anyPid };
}
Assert.assertArrayEquals(expectedCommand, checkProcessAliveCommand);
}
private void testInterval(long interval) throws IOException {