/**
* Return the level at which we are allowed to schedule containers. Given the thresholds
* indicating how much time passed before relaxing scheduling constraints.
*/
public synchronized NodeType getAllowedLocalityLevelByTime(
Priority priority, long nodeLocalityDelayMs, long rackLocalityDelayMs, long currentTimeMs) {
// if not being used, can schedule anywhere
if (nodeLocalityDelayMs < 0 || rackLocalityDelayMs < 0) {
return NodeType.OFF_SWITCH;
}
// default level is NODE_LOCAL
if (!allowedLocalityLevel.containsKey(priority)) {
allowedLocalityLevel.put(priority, NodeType.NODE_LOCAL);
return NodeType.NODE_LOCAL;
}
NodeType allowed = allowedLocalityLevel.get(priority);
// if level is already most liberal, we're done
if (allowed.equals(NodeType.OFF_SWITCH)) {
return NodeType.OFF_SWITCH;
}
// check waiting time
long waitTime = currentTimeMs;
if (lastScheduledContainer.containsKey(priority)) {
waitTime -= lastScheduledContainer.get(priority);
} else {
waitTime -= getStartTime();
}
long thresholdTime =
allowed.equals(NodeType.NODE_LOCAL) ? nodeLocalityDelayMs : rackLocalityDelayMs;
if (waitTime > thresholdTime) {
if (allowed.equals(NodeType.NODE_LOCAL)) {
allowedLocalityLevel.put(priority, NodeType.RACK_LOCAL);
resetSchedulingOpportunities(priority, currentTimeMs);
} else if (allowed.equals(NodeType.RACK_LOCAL)) {
allowedLocalityLevel.put(priority, NodeType.OFF_SWITCH);
resetSchedulingOpportunities(priority, currentTimeMs);
}
}
return allowedLocalityLevel.get(priority);
}
/**
* Return the level at which we are allowed to schedule containers, given the current size of the
* cluster and thresholds indicating how many nodes to fail at (as a fraction of cluster size)
* before relaxing scheduling constraints.
*/
public synchronized NodeType getAllowedLocalityLevel(
Priority priority, int numNodes, double nodeLocalityThreshold, double rackLocalityThreshold) {
// upper limit on threshold
if (nodeLocalityThreshold > 1.0) {
nodeLocalityThreshold = 1.0;
}
if (rackLocalityThreshold > 1.0) {
rackLocalityThreshold = 1.0;
}
// If delay scheduling is not being used, can schedule anywhere
if (nodeLocalityThreshold < 0.0 || rackLocalityThreshold < 0.0) {
return NodeType.OFF_SWITCH;
}
// Default level is NODE_LOCAL
if (!allowedLocalityLevel.containsKey(priority)) {
allowedLocalityLevel.put(priority, NodeType.NODE_LOCAL);
return NodeType.NODE_LOCAL;
}
NodeType allowed = allowedLocalityLevel.get(priority);
// If level is already most liberal, we're done
if (allowed.equals(NodeType.OFF_SWITCH)) return NodeType.OFF_SWITCH;
double threshold =
allowed.equals(NodeType.NODE_LOCAL) ? nodeLocalityThreshold : rackLocalityThreshold;
// Relax locality constraints once we've surpassed threshold.
if (getSchedulingOpportunities(priority) > (numNodes * threshold)) {
if (allowed.equals(NodeType.NODE_LOCAL)) {
allowedLocalityLevel.put(priority, NodeType.RACK_LOCAL);
resetSchedulingOpportunities(priority);
} else if (allowed.equals(NodeType.RACK_LOCAL)) {
allowedLocalityLevel.put(priority, NodeType.OFF_SWITCH);
resetSchedulingOpportunities(priority);
}
}
return allowedLocalityLevel.get(priority);
}
private Resource assignContainer(FSSchedulerNode node, boolean reserved) {
if (LOG.isDebugEnabled()) {
LOG.debug("Node offered to app: " + getName() + " reserved: " + reserved);
}
Collection<Priority> prioritiesToTry =
(reserved)
? Arrays.asList(node.getReservedContainer().getReservedPriority())
: getPriorities();
// For each priority, see if we can schedule a node local, rack local
// or off-switch request. Rack of off-switch requests may be delayed
// (not scheduled) in order to promote better locality.
synchronized (this) {
for (Priority priority : prioritiesToTry) {
if (getTotalRequiredResources(priority) <= 0 || !hasContainerForNode(priority, node)) {
continue;
}
addSchedulingOpportunity(priority);
// Check the AM resource usage for the leaf queue
if (getLiveContainers().size() == 0 && !getUnmanagedAM()) {
if (!getQueue().canRunAppAM(getAMResource())) {
return Resources.none();
}
}
ResourceRequest rackLocalRequest = getResourceRequest(priority, node.getRackName());
ResourceRequest localRequest = getResourceRequest(priority, node.getNodeName());
if (localRequest != null && !localRequest.getRelaxLocality()) {
LOG.warn("Relax locality off is not supported on local request: " + localRequest);
}
NodeType allowedLocality;
if (scheduler.isContinuousSchedulingEnabled()) {
allowedLocality =
getAllowedLocalityLevelByTime(
priority,
scheduler.getNodeLocalityDelayMs(),
scheduler.getRackLocalityDelayMs(),
scheduler.getClock().getTime());
} else {
allowedLocality =
getAllowedLocalityLevel(
priority,
scheduler.getNumClusterNodes(),
scheduler.getNodeLocalityThreshold(),
scheduler.getRackLocalityThreshold());
}
if (rackLocalRequest != null
&& rackLocalRequest.getNumContainers() != 0
&& localRequest != null
&& localRequest.getNumContainers() != 0) {
return assignContainer(node, localRequest, NodeType.NODE_LOCAL, reserved);
}
if (rackLocalRequest != null && !rackLocalRequest.getRelaxLocality()) {
continue;
}
if (rackLocalRequest != null
&& rackLocalRequest.getNumContainers() != 0
&& (allowedLocality.equals(NodeType.RACK_LOCAL)
|| allowedLocality.equals(NodeType.OFF_SWITCH))) {
return assignContainer(node, rackLocalRequest, NodeType.RACK_LOCAL, reserved);
}
ResourceRequest offSwitchRequest = getResourceRequest(priority, ResourceRequest.ANY);
if (offSwitchRequest != null && !offSwitchRequest.getRelaxLocality()) {
continue;
}
if (offSwitchRequest != null
&& offSwitchRequest.getNumContainers() != 0
&& allowedLocality.equals(NodeType.OFF_SWITCH)) {
return assignContainer(node, offSwitchRequest, NodeType.OFF_SWITCH, reserved);
}
}
}
return Resources.none();
}
public synchronized RMContainer allocate(
NodeType type,
FSSchedulerNode node,
Priority priority,
ResourceRequest request,
Container container) {
// Update allowed locality level
NodeType allowed = allowedLocalityLevel.get(priority);
if (allowed != null) {
if (allowed.equals(NodeType.OFF_SWITCH)
&& (type.equals(NodeType.NODE_LOCAL) || type.equals(NodeType.RACK_LOCAL))) {
this.resetAllowedLocalityLevel(priority, type);
} else if (allowed.equals(NodeType.RACK_LOCAL) && type.equals(NodeType.NODE_LOCAL)) {
this.resetAllowedLocalityLevel(priority, type);
}
}
// Required sanity check - AM can call 'allocate' to update resource
// request without locking the scheduler, hence we need to check
if (getTotalRequiredResources(priority) <= 0) {
return null;
}
// Create RMContainer
RMContainer rmContainer =
new RMContainerImpl(
container,
getApplicationAttemptId(),
node.getNodeID(),
appSchedulingInfo.getUser(),
rmContext);
// Add it to allContainers list.
newlyAllocatedContainers.add(rmContainer);
liveContainers.put(container.getId(), rmContainer);
// Update consumption and track allocations
List<ResourceRequest> resourceRequestList =
appSchedulingInfo.allocate(type, node, priority, request, container);
Resources.addTo(currentConsumption, container.getResource());
// Update resource requests related to "request" and store in RMContainer
((RMContainerImpl) rmContainer).setResourceRequests(resourceRequestList);
// Inform the container
rmContainer.handle(new RMContainerEvent(container.getId(), RMContainerEventType.START));
if (LOG.isDebugEnabled()) {
LOG.debug(
"allocate: applicationAttemptId="
+ container.getId().getApplicationAttemptId()
+ " container="
+ container.getId()
+ " host="
+ container.getNodeId().getHost()
+ " type="
+ type);
}
RMAuditLogger.logSuccess(
getUser(),
AuditConstants.ALLOC_CONTAINER,
"SchedulerApp",
getApplicationId(),
container.getId());
return rmContainer;
}
private Resource assignContainer(
FSSchedulerNode node, boolean reserved, TransactionState transactionState) {
if (LOG.isDebugEnabled()) {
LOG.debug("Node offered to app: " + getName() + " reserved: " + reserved);
}
if (reserved) {
RMContainer rmContainer = node.getReservedContainer();
Priority priority = rmContainer.getReservedPriority();
// Make sure the application still needs requests at this priority
if (app.getTotalRequiredResources(priority) == 0) {
unreserve(priority, node);
return Resources.none();
}
}
Collection<Priority> prioritiesToTry =
(reserved)
? Arrays.asList(node.getReservedContainer().getReservedPriority())
: app.getPriorities();
// For each priority, see if we can schedule a node local, rack local
// or off-switch request. Rack of off-switch requests may be delayed
// (not scheduled) in order to promote better locality.
synchronized (app) {
for (Priority priority : prioritiesToTry) {
if (app.getTotalRequiredResources(priority) <= 0 || !hasContainerForNode(priority, node)) {
continue;
}
app.addSchedulingOpportunity(priority);
ResourceRequest rackLocalRequest = app.getResourceRequest(priority, node.getRackName());
ResourceRequest localRequest = app.getResourceRequest(priority, node.getNodeName());
if (localRequest != null && !localRequest.getRelaxLocality()) {
LOG.warn("Relax locality off is not supported on local request: " + localRequest);
}
NodeType allowedLocality;
if (scheduler.isContinuousSchedulingEnabled()) {
allowedLocality =
app.getAllowedLocalityLevelByTime(
priority,
scheduler.getNodeLocalityDelayMs(),
scheduler.getRackLocalityDelayMs(),
scheduler.getClock().getTime());
} else {
allowedLocality =
app.getAllowedLocalityLevel(
priority,
scheduler.getNumClusterNodes(),
scheduler.getNodeLocalityThreshold(),
scheduler.getRackLocalityThreshold());
}
if (rackLocalRequest != null
&& rackLocalRequest.getNumContainers() != 0
&& localRequest != null
&& localRequest.getNumContainers() != 0) {
return assignContainer(
node, priority, localRequest, NodeType.NODE_LOCAL, reserved, transactionState);
}
if (rackLocalRequest != null && !rackLocalRequest.getRelaxLocality()) {
continue;
}
if (rackLocalRequest != null
&& rackLocalRequest.getNumContainers() != 0
&& (allowedLocality.equals(NodeType.RACK_LOCAL)
|| allowedLocality.equals(NodeType.OFF_SWITCH))) {
return assignContainer(
node, priority, rackLocalRequest, NodeType.RACK_LOCAL, reserved, transactionState);
}
ResourceRequest offSwitchRequest = app.getResourceRequest(priority, ResourceRequest.ANY);
if (offSwitchRequest != null && !offSwitchRequest.getRelaxLocality()) {
continue;
}
if (offSwitchRequest != null
&& offSwitchRequest.getNumContainers() != 0
&& allowedLocality.equals(NodeType.OFF_SWITCH)) {
return assignContainer(
node, priority, offSwitchRequest, NodeType.OFF_SWITCH, reserved, transactionState);
}
}
}
return Resources.none();
}
