Hendrik/xgboost
1
0
Fork 0
Code Issues Pull Requests Packages Projects Releases Wiki Activity

OK

Browse Source
This commit is contained in:
tqchen 2015-03-10 10:51:39 -07:00
parent 091634b259
commit 014c86603d
1 changed files with 10 additions and 2 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

12
yarn/src/org/apache/hadoop/yarn/rabit/ApplicationMaster.java
View File

@ -34,6 +34,7 @@ import org.apache.hadoop.yarn.api.records.NodeReport;
import org.apache.hadoop.yarn.client.api.AMRMClient.ContainerRequest; import org.apache.hadoop.yarn.client.api.AMRMClient.ContainerRequest;
import org.apache.hadoop.yarn.client.api.async.NMClientAsync; import org.apache.hadoop.yarn.client.api.async.NMClientAsync;
import org.apache.hadoop.yarn.client.api.async.AMRMClientAsync; import org.apache.hadoop.yarn.client.api.async.AMRMClientAsync;
import org.apache.hadoop.security.UserGroupInformation;
/** /**
* application master for allocating resources of rabit client * application master for allocating resources of rabit client
@ -61,6 +62,8 @@ public class ApplicationMaster {
// command to launch // command to launch
private String command = ""; private String command = "";
// username
private String userName = "";
// application tracker hostname // application tracker hostname
private String appHostName = ""; private String appHostName = "";
// tracker URL to do // tracker URL to do
@ -128,6 +131,8 @@ public class ApplicationMaster {
*/ */
private void initArgs(String args[]) throws IOException { private void initArgs(String args[]) throws IOException {
LOG.info("Invoke initArgs"); LOG.info("Invoke initArgs");
// get user name
userName = UserGroupInformation.getCurrentUser().getShortUserName();
// cached maps // cached maps
Map<String, Path> cacheFiles = new java.util.HashMap<String, Path>(); Map<String, Path> cacheFiles = new java.util.HashMap<String, Path>();
for (int i = 0; i < args.length; ++i) { for (int i = 0; i < args.length; ++i) {
@ -272,7 +277,6 @@ public class ApplicationMaster {
+ ApplicationConstants.LOG_DIR_EXPANSION_VAR + "/stdout" + ApplicationConstants.LOG_DIR_EXPANSION_VAR + "/stdout"
+ " 2>" + ApplicationConstants.LOG_DIR_EXPANSION_VAR + " 2>" + ApplicationConstants.LOG_DIR_EXPANSION_VAR
+ "/stderr"; + "/stderr";
LOG.info(cmd);
ctx.setCommands(Collections.singletonList(cmd)); ctx.setCommands(Collections.singletonList(cmd));
LOG.info(workerResources); LOG.info(workerResources);
ctx.setLocalResources(this.workerResources); ctx.setLocalResources(this.workerResources);
@ -376,8 +380,12 @@ public class ApplicationMaster {
Collection<TaskRecord> tasks = new java.util.LinkedList<TaskRecord>(); Collection<TaskRecord> tasks = new java.util.LinkedList<TaskRecord>();
for (ContainerId cid : failed) { for (ContainerId cid : failed) {
TaskRecord r = runningTasks.remove(cid); TaskRecord r = runningTasks.remove(cid);
if (r == null) if (r == null) {
continue; continue;
}
LOG.info("Task " + r.taskId + "failed on " + r.container.getId() + ". See LOG at : " +
String.format("http://%s/node/containerlogs/%s/" + userName,
r.container.getNodeHttpAddress(), r.container.getId()));
r.attemptCounter += 1; r.attemptCounter += 1;
r.container = null; r.container = null;
tasks.add(r); tasks.add(r);