private static void publishContainerEndEvent(
TimelineClient timelineClient, ContainerStatus container) throws IOException, YarnException {
TimelineEntity entity = new TimelineEntity();
entity.setEntityId(container.getContainerId().toString());
entity.setEntityType(DSEntity.DS_CONTAINER.toString());
entity.addPrimaryFilter("user", UserGroupInformation.getCurrentUser().toString());
TimelineEvent event = new TimelineEvent();
event.setTimestamp(System.currentTimeMillis());
event.setEventType(DSEvent.DS_CONTAINER_END.toString());
event.addEventInfo("State", container.getState().name());
event.addEventInfo("Exit Status", container.getExitStatus());
entity.addEvent(event);
timelineClient.putEntities(entity);
}
private NodeStatus getNodeStatus() {
NodeStatus nodeStatus = recordFactory.newRecordInstance(NodeStatus.class);
nodeStatus.setNodeId(this.nodeId);
int numActiveContainers = 0;
List<ContainerStatus> containersStatuses = new ArrayList<ContainerStatus>();
for (Iterator<Entry<ContainerId, Container>> i =
this.context.getContainers().entrySet().iterator();
i.hasNext(); ) {
Entry<ContainerId, Container> e = i.next();
ContainerId containerId = e.getKey();
Container container = e.getValue();
// Clone the container to send it to the RM
org.apache.hadoop.yarn.api.records.ContainerStatus containerStatus =
container.cloneAndGetContainerStatus();
containersStatuses.add(containerStatus);
++numActiveContainers;
LOG.info("Sending out status for container: " + containerStatus);
if (containerStatus.getState() == ContainerState.COMPLETE) {
// Remove
i.remove();
LOG.info("Removed completed container " + containerId);
}
}
nodeStatus.setContainersStatuses(containersStatuses);
LOG.debug(this.nodeId + " sending out status for " + numActiveContainers + " containers");
NodeHealthStatus nodeHealthStatus = this.context.getNodeHealthStatus();
if (this.healthChecker != null) {
this.healthChecker.setHealthStatus(nodeHealthStatus);
}
LOG.debug(
"Node's health-status : "
+ nodeHealthStatus.getIsNodeHealthy()
+ ", "
+ nodeHealthStatus.getHealthReport());
nodeStatus.setNodeHealthStatus(nodeHealthStatus);
return nodeStatus;
}
private ImmutableMap.Builder<String, String> buildContainerStatusEventMetadata(
ContainerStatus containerStatus) {
ImmutableMap.Builder<String, String> eventMetadataBuilder = new ImmutableMap.Builder<>();
eventMetadataBuilder.put(
GobblinYarnMetricTagNames.CONTAINER_ID, containerStatus.getContainerId().toString());
eventMetadataBuilder.put(
GobblinYarnEventConstants.EventMetadata.CONTAINER_STATUS_CONTAINER_STATE,
containerStatus.getState().toString());
if (ContainerExitStatus.INVALID != containerStatus.getExitStatus()) {
eventMetadataBuilder.put(
GobblinYarnEventConstants.EventMetadata.CONTAINER_STATUS_EXIT_STATUS,
containerStatus.getExitStatus() + "");
}
if (!Strings.isNullOrEmpty(containerStatus.getDiagnostics())) {
eventMetadataBuilder.put(
GobblinYarnEventConstants.EventMetadata.CONTAINER_STATUS_EXIT_DIAGNOSTICS,
containerStatus.getDiagnostics());
}
return eventMetadataBuilder;
}
@SuppressWarnings("unchecked")
@Override
public void onContainersCompleted(List<ContainerStatus> completedContainers) {
LOG.info(
"Got response from RM for container ask, completedCnt=" + completedContainers.size());
for (ContainerStatus containerStatus : completedContainers) {
LOG.info(
"Got container status for containerID="
+ containerStatus.getContainerId()
+ ", state="
+ containerStatus.getState()
+ ", exitStatus="
+ containerStatus.getExitStatus()
+ ", diagnostics="
+ containerStatus.getDiagnostics());
// non complete containers should not be here
assert (containerStatus.getState() == ContainerState.COMPLETE);
// increment counters for completed/failed containers
int exitStatus = containerStatus.getExitStatus();
if (0 != exitStatus) {
// container failed
if (ContainerExitStatus.ABORTED != exitStatus) {
// shell script failed
// counts as completed
numCompletedContainers.incrementAndGet();
numFailedContainers.incrementAndGet();
} else {
// container was killed by framework, possibly preempted
// we should re-try as the container was lost for some reason
numAllocatedContainers.decrementAndGet();
numRequestedContainers.decrementAndGet();
// we do not need to release the container as it would be done
// by the RM
}
} else {
// nothing to do
// container completed successfully
numCompletedContainers.incrementAndGet();
LOG.info(
"Container completed successfully."
+ ", containerId="
+ containerStatus.getContainerId());
}
try {
publishContainerEndEvent(timelineClient, containerStatus);
} catch (Exception e) {
LOG.error(
"Container start event could not be pulished for "
+ containerStatus.getContainerId().toString(),
e);
}
}
// ask for more containers if any failed
int askCount = numTotalContainers - numRequestedContainers.get();
numRequestedContainers.addAndGet(askCount);
if (askCount > 0) {
for (int i = 0; i < askCount; ++i) {
ContainerRequest containerAsk = setupContainerAskForRM();
amRMClient.addContainerRequest(containerAsk);
}
}
if (numCompletedContainers.get() == numTotalContainers) {
done = true;
}
}