cameronlee314 commented on a change in pull request #1027: SAMZA-2046:
Startpoint fan out implementation
URL: https://github.com/apache/samza/pull/1027#discussion_r283520468
##########
File path:
samza-core/src/main/java/org/apache/samza/startpoint/StartpointManager.java
##########
@@ -173,71 +226,151 @@ public void deleteStartpoint(SystemStreamPartition ssp,
TaskName taskName) {
Preconditions.checkState(!stopped, "Underlying metadata store not
available");
Preconditions.checkNotNull(ssp, "SystemStreamPartition cannot be null");
- metadataStore.delete(toStoreKey(ssp, taskName));
+ readWriteStore.delete(toReadWriteStoreKey(ssp, taskName));
}
/**
- * For {@link Startpoint}s keyed only by {@link SystemStreamPartition}, this
method re-maps the Startpoints from
- * SystemStreamPartition to SystemStreamPartition+{@link TaskName} for all
tasks provided by the {@link JobModel}
+ * The Startpoints that are written to with {@link
#writeStartpoint(SystemStreamPartition, Startpoint)} and with
+ * {@link #writeStartpoint(SystemStreamPartition, TaskName, Startpoint)} are
moved from a "read-write" namespace
+ * to a "fan out" namespace.
* This method is not atomic or thread-safe. The intent is for the Samza
Processor's coordinator to use this
* method to assign the Startpoints to the appropriate tasks.
- * @param jobModel The {@link JobModel} is used to determine which {@link
TaskName} each {@link SystemStreamPartition} maps to.
- * @return The list of {@link SystemStreamPartition}s that were fanned out
to SystemStreamPartition+TaskName.
+ * @param taskToSSPs Determines which {@link TaskName} each {@link
SystemStreamPartition} maps to.
+ * @return The set of active {@link TaskName}s that were fanned out to.
*/
- public Set<SystemStreamPartition> fanOutStartpointsToTasks(JobModel
jobModel) {
+ public Set<TaskName> fanOut(Map<TaskName, Set<SystemStreamPartition>>
taskToSSPs) {
Preconditions.checkState(!stopped, "Underlying metadata store not
available");
- Preconditions.checkNotNull(jobModel, "JobModel cannot be null");
-
- HashSet<SystemStreamPartition> sspsToDelete = new HashSet<>();
-
- // Inspect the job model for TaskName-to-SSPs mapping and re-map
startpoints from SSP-only keys to SSP+TaskName keys.
- for (ContainerModel containerModel: jobModel.getContainers().values()) {
- for (TaskModel taskModel : containerModel.getTasks().values()) {
- TaskName taskName = taskModel.getTaskName();
- for (SystemStreamPartition ssp :
taskModel.getSystemStreamPartitions()) {
- Startpoint startpoint = readStartpoint(ssp); // Read SSP-only key
- if (startpoint == null) {
- LOG.debug("No Startpoint for SSP: {} in task: {}", ssp, taskName);
- continue;
- }
-
- LOG.info("Grouping Startpoint keyed on SSP: {} to tasks determined
by the job model.", ssp);
- Startpoint startpointForTask = readStartpoint(ssp, taskName);
- if (startpointForTask == null ||
startpointForTask.getCreationTimestamp() < startpoint.getCreationTimestamp()) {
- writeStartpoint(ssp, taskName, startpoint);
- sspsToDelete.add(ssp); // Mark for deletion
- LOG.info("Startpoint for SSP: {} remapped with task: {}.", ssp,
taskName);
- } else {
- LOG.info("Startpoint for SSP: {} and task: {} already exists and
will not be overwritten.", ssp, taskName);
- }
+ Preconditions.checkArgument(MapUtils.isNotEmpty(taskToSSPs), "taskToSSPs
cannot be null or empty");
+
+ // construct fan out with the existing readWriteStore entries and mark the
entries for deletion after fan out
+ Instant now = Instant.now();
+ HashMultimap<SystemStreamPartition, TaskName> deleteKeys =
HashMultimap.create();
+ HashMap<TaskName, StartpointFanOut> fanOuts = new HashMap<>();
+ for (TaskName taskName : taskToSSPs.keySet()) {
+ Set<SystemStreamPartition> ssps = taskToSSPs.get(taskName);
+ if (CollectionUtils.isEmpty(ssps)) {
+ LOG.warn("No SSPs are mapped to taskName: {}", taskName.getTaskName());
+ continue;
+ }
+ for (SystemStreamPartition ssp : ssps) {
+ Startpoint startpoint = readStartpoint(ssp); // Read SSP-only key
+ if (startpoint != null) {
+ deleteKeys.put(ssp, null);
+ }
+
+ Startpoint startpointForTask = readStartpoint(ssp, taskName); // Read
SSP+taskName key
+ if (startpointForTask != null) {
+ deleteKeys.put(ssp, taskName);
+ }
+
+ Startpoint startpointWithPrecedence =
resolveStartpointPrecendence(startpoint, startpointForTask);
+ if (startpointWithPrecedence == null) {
+ continue;
+ }
+
+ fanOuts.putIfAbsent(taskName, new StartpointFanOut(now, new
HashMap<>()));
+ fanOuts.get(taskName).getFanOuts().put(ssp, startpointWithPrecedence);
+ }
+ }
+
+ if (fanOuts.isEmpty()) {
+ LOG.debug("No fan outs created.");
+ return ImmutableSet.of();
+ }
+
+ LOG.info("Fanning out to {} tasks", fanOuts.size());
+
+ StartpointFanOut.StartpointFanOutSerde startpointFanOutSerde = new
StartpointFanOut.StartpointFanOutSerde();
+ for (TaskName taskName : fanOuts.keySet()) {
+ String fanOutKey = toFanOutStoreKey(taskName);
+ StartpointFanOut newFanOut = fanOuts.get(taskName);
+ byte[] fanOutFromStoreInBytes = fanOutStore.get(fanOutKey);
+
+ if (ArrayUtils.isNotEmpty(fanOutFromStoreInBytes)) {
+ // Merge new fan out with existing fan out. This is not a typical
scenario but it can happen when the job
+ // restarts before all startpoints are committed by the first
checkpoint commit and new startpoints were written
+ // to the readWriteStore after the previous fan out.
+ StartpointFanOut existingFanOut =
startpointFanOutSerde.fromBytes(fanOutFromStoreInBytes);
+ existingFanOut.getFanOuts().putAll(newFanOut.getFanOuts());
+ newFanOut = existingFanOut;
+ }
+
+ fanOutStore.put(fanOutKey, startpointFanOutSerde.toBytes(newFanOut));
+ // Delete from readWriteStore after fan out.
+ for (SystemStreamPartition ssp : newFanOut.getFanOuts().keySet()) {
+ if (deleteKeys.remove(ssp, taskName)) {
+ deleteStartpoint(ssp, taskName);
+ }
+ if (deleteKeys.remove(ssp, null)) {
+ deleteStartpoint(ssp);
}
}
}
- // Delete SSP-only keys
- sspsToDelete.forEach(ssp -> {
- deleteStartpoint(ssp);
- LOG.info("All Startpoints for SSP: {} have been grouped to the
appropriate tasks and the SSP was deleted.");
- });
+ return ImmutableSet.copyOf(fanOuts.keySet());
+ }
+
+ /**
+ * Read the fanned out {@link Startpoint}s for the given {@link TaskName}
+ * @param taskName to read the fan out Startpoints for
+ * @return fanned out Startpoints
+ */
+ public Map<SystemStreamPartition, Startpoint> getFanOutForTask(TaskName
taskName) {
Review comment:
There seems to be 3 different "clients" for the startpoints stores:
1) startpoint tool (write to non-fanout store)
2) job coordinator (read/write to non-fanout store, write to fanout store)
3) processing containers (read/write to fanout store)
Would it be good to separate out these categories into different classes?
Then we wouldn't need to construct the full `StartpointManager` for everything.
One main benefit I see from this is that it will be easier to construct
dependencies for each stage. One non-ideal pattern in Samza that I have
observed is that the same object is used for different flows, so all of the
dependencies for that object are needed, even if that flow doesn't actually use
the dependency. Do you think it would be possible/helpful to avoid that pattern
here?
----------------------------------------------------------------
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.
For queries about this service, please contact Infrastructure at:
[email protected]
With regards,
Apache Git Services
