External Storage Integration: NexusWorker #3018
There are no files selected for viewing
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
| Original file line number | Diff line number | Diff line change |
|---|---|---|
|
|
@@ -4,6 +4,7 @@ | |
| import static io.temporal.serviceclient.MetricsTag.TASK_FAILURE_TYPE; | ||
|
|
||
| import com.google.protobuf.ByteString; | ||
| import com.google.protobuf.Message; | ||
| import com.uber.m3.tally.Scope; | ||
| import com.uber.m3.tally.Stopwatch; | ||
| import com.uber.m3.util.Duration; | ||
|
|
@@ -16,7 +17,9 @@ | |
| import io.temporal.common.converter.DataConverter; | ||
| import io.temporal.internal.common.NexusUtil; | ||
| import io.temporal.internal.common.ProtobufTimeUtils; | ||
| import io.temporal.internal.concurrent.structured.CancelSource; | ||
| import io.temporal.internal.logging.LoggerTag; | ||
| import io.temporal.internal.payload.storage.ExternalStorageRunner; | ||
| import io.temporal.internal.retryer.GrpcRetryer; | ||
| import io.temporal.serviceclient.MetricsTag; | ||
| import io.temporal.serviceclient.WorkflowServiceStubs; | ||
|
|
@@ -27,6 +30,7 @@ | |
| import io.temporal.worker.tuning.PollerBehaviorAutoscaling; | ||
| import java.util.Collections; | ||
| import java.util.Objects; | ||
| import java.util.concurrent.CancellationException; | ||
| import java.util.concurrent.CompletableFuture; | ||
| import java.util.concurrent.TimeUnit; | ||
| import java.util.concurrent.TimeoutException; | ||
|
|
@@ -53,6 +57,9 @@ final class NexusWorker implements SuspendableWorker { | |
| private final GrpcRetryer.GrpcRetryerOptions replyGrpcRetryerOptions; | ||
| private final TrackingSlotSupplier<NexusSlotInfo> slotSupplier; | ||
| private final NamespaceCapabilities namespaceCapabilities; | ||
|
|
||
| final CancelSource<CancellationException> storageCancellation = | ||
| new CancelSource<>(() -> new CancellationException("Worker shutdown")); | ||
| private final boolean forceOldFailureFormat; | ||
| private final boolean workerCommandsTaskQueue; | ||
| private final TaskCounter taskCounter = new TaskCounter(); | ||
|
|
@@ -182,6 +189,9 @@ public boolean start() { | |
|
|
||
| @Override | ||
| public CompletableFuture<Void> shutdown(ShutdownManager shutdownManager, boolean interruptTasks) { | ||
| if (interruptTasks) { | ||
| storageCancellation.cancel(); | ||
| } | ||
| String supplierName = this + "#executorSlots"; | ||
| return poller | ||
| .shutdown(shutdownManager, interruptTasks) | ||
|
|
@@ -274,6 +284,12 @@ public String toString() { | |
| options.getIdentity(), namespace, taskQueue); | ||
| } | ||
|
|
||
| private static final class ExternalStorageTaskFailure extends RuntimeException { | ||
| ExternalStorageTaskFailure(String message, Throwable cause) { | ||
| super(message, cause); | ||
| } | ||
| } | ||
|
|
||
| private class TaskHandlerImpl implements PollTaskExecutor.TaskHandler<NexusTask> { | ||
|
|
||
| final NexusTaskHandler handler; | ||
|
|
@@ -304,28 +320,38 @@ private String getNexusTaskOperation(PollNexusTaskQueueResponseOrBuilder pollRes | |
|
|
||
| @Override | ||
| public void handle(NexusTask task) { | ||
| boolean taskFailed = false; | ||
| try { | ||
| PollNexusTaskQueueResponseOrBuilder pollResponse = task.getResponse(); | ||
| // Extract service and operation from the request and set them as MDC and metrics | ||
| // scope tags. If the request does not have a service or operation, do not set the tags. | ||
| // If we don't know how to handle the task, we will fail the task further down the line. | ||
| Scope metricsScope = workerMetricsScope; | ||
| String service = getNexusTaskService(pollResponse); | ||
| if (!service.isEmpty()) { | ||
| MDC.put(LoggerTag.NEXUS_SERVICE, service); | ||
| metricsScope = metricsScope.tagged(ImmutableMap.of(MetricsTag.NEXUS_SERVICE, service)); | ||
| } | ||
| String operation = getNexusTaskOperation(pollResponse); | ||
| if (!operation.isEmpty()) { | ||
| MDC.put(LoggerTag.NEXUS_OPERATION, operation); | ||
| metricsScope = | ||
| metricsScope.tagged(ImmutableMap.of(MetricsTag.NEXUS_OPERATION, operation)); | ||
| } | ||
| slotSupplier.markSlotUsed( | ||
| new NexusSlotInfo( | ||
| service, operation, taskQueue, options.getIdentity(), options.getBuildId()), | ||
| task.getPermit()); | ||
|
Contributor
There was a problem hiding this comment. This code and above used to be outside of the try block. Not sure it needs to be inside. It might incidentally fix some existing problems. But its not the focus of this PR. If it doesn't need to be in the try to enable external storage, then let's move it back. If it is fixing something separately, then open a different PR for it. |
||
|
|
||
| try { | ||
| task = retrieveInboundPayloads(task); | ||
| } catch (Throwable e) { | ||
| taskFailed = true; | ||
| sendStorageFailure( | ||
|
Contributor
There was a problem hiding this comment. This is going to log "External storage failed for a nexus task" but external storage is not configured, so the log is misleading. |
||
| pollResponse.getTaskToken(), supportsTemporalFailure(pollResponse), metricsScope, e); | ||
| return; | ||
| } | ||
|
|
||
| taskFailed = handleNexusTask(task, metricsScope); | ||
| } catch (Throwable e) { | ||
| taskFailed = true; | ||
|
|
@@ -416,14 +442,11 @@ private boolean handleNexusTask(NexusTask task, Scope metricsScope) { | |
| } | ||
|
|
||
| try { | ||
| sendReply(taskToken, supportsTemporalFailure(pollResponse), result, metricsScope); | ||
| } catch (ExternalStorageTaskFailure e) { | ||
| sendStorageFailure( | ||
| taskToken, supportsTemporalFailure(pollResponse), metricsScope, e.getCause()); | ||
| return true; | ||
| } catch (Exception e) { | ||
| logExceptionDuringResultReporting(e, pollResponse, result); | ||
| throw e; | ||
|
|
@@ -484,13 +507,14 @@ private void sendReply( | |
| if (!supportTemporalFailure && taskResponse.getStartOperation().hasFailure()) { | ||
| taskResponse = getResponseForOldServer(taskResponse); | ||
| } | ||
| RespondNexusTaskCompletedRequest.Builder requestBuilder = | ||
| RespondNexusTaskCompletedRequest.newBuilder() | ||
| .setTaskToken(taskToken) | ||
| .setIdentity(options.getIdentity()) | ||
| .setNamespace(namespace) | ||
| .setResponse(taskResponse); | ||
| storeOutbound(requestBuilder); | ||
| RespondNexusTaskCompletedRequest request = requestBuilder.build(); | ||
|
|
||
| grpcRetryer.retry( | ||
| () -> | ||
|
|
@@ -512,17 +536,71 @@ private void sendReply( | |
| } else { | ||
| request.setError(NexusUtil.handlerErrorToNexusError(handlerException, dataConverter)); | ||
| } | ||
| storeOutbound(request); | ||
| RespondNexusTaskFailedRequest failedRequest = request.build(); | ||
| grpcRetryer.retry( | ||
| () -> | ||
| service | ||
| .blockingStub() | ||
| .withOption(METRICS_TAGS_CALL_OPTIONS_KEY, metricsScope) | ||
| .respondNexusTaskFailed(failedRequest), | ||
| replyGrpcRetryerOptions); | ||
| } else { | ||
| throw new IllegalArgumentException("[BUG] Either response or failure must be set"); | ||
| } | ||
| } | ||
| } | ||
|
|
||
| private boolean supportsTemporalFailure(PollNexusTaskQueueResponseOrBuilder pollResponse) { | ||
| return !forceOldFailureFormat | ||
| && pollResponse.getRequest().getCapabilities().getTemporalFailureResponses(); | ||
| } | ||
|
|
||
| private void sendStorageFailure( | ||
| ByteString taskToken, boolean supportTemporalFailure, Scope metricsScope, Throwable e) { | ||
| log.warn("External storage failed for a nexus task", e); | ||
| metricsScope | ||
| .tagged( | ||
| Collections.singletonMap( | ||
| TASK_FAILURE_TYPE, MetricsTag.TASK_FAILURE_VALUE_HANDLER_ERROR_INTERNAL)) | ||
| .counter(MetricsType.NEXUS_EXEC_FAILED_COUNTER) | ||
| .inc(1); | ||
| HandlerException handlerException = | ||
| new HandlerException(HandlerException.ErrorType.INTERNAL, "External storage failed", e); | ||
| sendReply( | ||
|
Contributor
There was a problem hiding this comment. I think this needs to handle |
||
| taskToken, | ||
| supportTemporalFailure, | ||
| new NexusTaskHandler.Result(handlerException), | ||
| metricsScope); | ||
| } | ||
|
|
||
| private NexusTask retrieveInboundPayloads(NexusTask task) { | ||
| ExternalStorageRunner externalStorageRunner = options.getExternalStorageRunner(); | ||
| PollNexusTaskQueueResponseOrBuilder response = task.getResponse(); | ||
| PollNexusTaskQueueResponse built = | ||
| response instanceof PollNexusTaskQueueResponse | ||
| ? (PollNexusTaskQueueResponse) response | ||
| : ((PollNexusTaskQueueResponse.Builder) response).build(); | ||
| if (externalStorageRunner == null) { | ||
| ExternalStorageRunner.throwIfContainsReference(built); | ||
| return task; | ||
| } | ||
| return new NexusTask( | ||
| externalStorageRunner.retrieve(built, storageCancellation.token()), | ||
| task.getPermit(), | ||
| task.getCompletionCallback()); | ||
| } | ||
|
|
||
| private void storeOutbound(Message.Builder builder) { | ||
|
Contributor
There was a problem hiding this comment. Why does this
Contributor
Author
There was a problem hiding this comment. Some callers of |
||
| ExternalStorageRunner externalStorageRunner = options.getExternalStorageRunner(); | ||
| if (externalStorageRunner == null) { | ||
| return; | ||
| } | ||
| try { | ||
| externalStorageRunner.store(builder, null, null, storageCancellation.token()); | ||
| } catch (Throwable e) { | ||
| throw new ExternalStorageTaskFailure("External storage store failed", e); | ||
| } | ||
| } | ||
| } | ||
| } | ||
Oops, something went wrong.
Oops, something went wrong.
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
Should this
setExternalStorage(null)?