Github user HeartSaVioR commented on a diff in the pull request:
https://github.com/apache/storm/pull/2241#discussion_r129515502
--- Diff: storm-client/src/jvm/org/apache/storm/daemon/Task.java ---
@@ -122,28 +130,33 @@ public Task(Executor executor, Integer taskId) throws
IOException {
return new ArrayList<>(0);
}
+
public List<Integer> getOutgoingTasks(String stream, List<Object>
values) {
if (debug) {
LOG.info("Emitting Tuple: taskId={} componentId={} stream={}
values={}", taskId, componentId, stream, values);
}
- List<Integer> outTasks = new ArrayList<>();
- if (!streamComponentToGrouper.containsKey(stream)) {
- throw new IllegalArgumentException("Unknown stream ID: " +
stream);
- }
- if (null != streamComponentToGrouper.get(stream)) {
- // null value for __system
- for (LoadAwareCustomStreamGrouping grouper :
streamComponentToGrouper.get(stream).values()) {
+ ArrayList<Integer> outTasks = new ArrayList<>();
+
+ // TODO: PERF: expensive hashtable lookup in critical path
+ ArrayList<LoadAwareCustomStreamGrouping> groupers =
streamToGroupers.get(stream);
+ if (null != groupers) {
+ for (int i=0; i<groupers.size(); ++i) {
+ LoadAwareCustomStreamGrouping grouper = groupers.get(i);
if (grouper == GrouperFactory.DIRECT) {
throw new IllegalArgumentException("Cannot do regular
emit to direct stream");
}
List<Integer> compTasks = grouper.chooseTasks(taskId,
values, loadMapping);
- outTasks.addAll(compTasks);
+ outTasks.addAll(compTasks); // TODO: PERF: this is a
perf hit
--- End diff --
I'm curious why this line (not above line) becomes performance hit, since
compTasks would be a small value for most of groupers. Do you know why? Is it
related to memory reallocation for outTasks?
---
If your project is set up for it, you can reply to this email and have your
reply appear on GitHub as well. If your project does not have this feature
enabled and wishes so, or if the feature is enabled but not working, please
contact infrastructure at infrastruct...@apache.org or file a JIRA ticket
with INFRA.
---
