gortiz commented on code in PR #15445:
URL: https://github.com/apache/pinot/pull/15445#discussion_r2068157362
##########
pinot-query-runtime/src/main/java/org/apache/pinot/query/service/server/QueryServer.java:
##########
@@ -159,37 +164,49 @@ public void submit(Worker.QueryRequest request,
StreamObserver<Worker.QueryRespo
return;
}
- try (QueryThreadContext.CloseableContext queryTlClosable =
QueryThreadContext.openFromRequestMetadata(reqMetadata);
- QueryThreadContext.CloseableContext mseTlCloseable =
MseWorkerThreadContext.open()) {
+ try (QueryThreadContext.CloseableContext qClosable =
QueryThreadContext.openFromRequestMetadata(reqMetadata);
+ QueryThreadContext.CloseableContext mseCloseable =
MseWorkerThreadContext.open()) {
+
long requestId = QueryThreadContext.getRequestId();
QueryThreadContext.setQueryEngine("mse");
- Tracing.ThreadAccountantOps.setupRunner(Long.toString(requestId),
ThreadExecutionContext.TaskType.MSE);
- ThreadExecutionContext parentContext =
Tracing.getThreadAccountant().getThreadExecutionContext();
- try {
- forEachStage(request,
- (stagePlan, workerMetadata) -> {
+ // Submit the stage for each worker
+ List<CompletableFuture<List<Object>>> futures =
forEachStageAndWorker(request,
+ (stagePlan, workerMetadata) -> {
+ Tracing.ThreadAccountantOps.setupRunner(Long.toString(requestId),
ThreadExecutionContext.TaskType.MSE);
+ ThreadExecutionContext parentContext =
Tracing.getThreadAccountant().getThreadExecutionContext();
+
+ try {
_queryRunner.processQuery(workerMetadata, stagePlan,
reqMetadata, parentContext);
- return null;
- },
- (ignored) -> {
- });
- } catch (ExecutionException | InterruptedException | TimeoutException |
RuntimeException e) {
- LOGGER.error("Caught exception while submitting request: {}",
requestId, e);
- String errorMsg = "Caught exception while submitting request: " +
e.getMessage();
- responseObserver.onNext(Worker.QueryResponse.newBuilder()
-
.putMetadata(CommonConstants.Query.Response.ServerResponseStatus.STATUS_ERROR,
errorMsg)
- .build());
- responseObserver.onCompleted();
- return;
- } finally {
- Tracing.ThreadAccountantOps.clear();
- }
- responseObserver.onNext(
- Worker.QueryResponse.newBuilder()
-
.putMetadata(CommonConstants.Query.Response.ServerResponseStatus.STATUS_OK, "")
+ } finally {
+ Tracing.ThreadAccountantOps.clear();
+ }
+ return null;
+ });
+
+ // A completable future that will finish when all submit task finish or
on timoeut
+ CompletableFuture<Void> allCompleted =
CompletableFuture.allOf(futures.toArray(new CompletableFuture[0]))
+ .orTimeout(QueryThreadContext.getDeadlineMs(),
TimeUnit.MILLISECONDS);
+ // When this future completes, notify the broker.
+ allCompleted.handle((result, error) -> {
Review Comment:
AFAIK they are similar in this case. The main difference between handle and
whenComplete, is that handle can be used to _recover_ from the exception. But
given we don't use the result for anything, both should be equivalent. I can
change it to whenComplete if you find it easier to read
--
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.
To unsubscribe, e-mail: [email protected]
For queries about this service, please contact Infrastructure at:
[email protected]
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
