Skip to content

[Backport 3.0] Monitoring: Accurately capture predict failures (#4525) #4840

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Open
pyek-bot wants to merge 1 commit into
base: 3.0
Choose a base branch
from
+42 −2
Open

[Backport 3.0] Monitoring: Accurately capture predict failures (#4525) #4840

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
1 commit
Select commit
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
22 changes: 20 additions & 2 deletions plugin/src/main/java/org/opensearch/ml/task/MLPredictTaskRunner.java
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -461,7 +461,7 @@ private void runPredict(
mlModelManager.trackPredictDuration(modelId, startTime);
internalListener.onResponse(output);
}
}, e -> handlePredictFailure(mlTask, internalListener, e, false, modelId, actionName));
}, e -> handlePredictFailure(mlTask, internalListener, e, shouldTrackRemoteFailure(e), modelId, actionName));
predictor.asyncPredict(mlInput, trackPredictDurationListener);
} else {
MLOutput output = mlModelManager.trackPredictDuration(modelId, () -> predictor.predict(mlInput));
Expand All @@ -478,7 +478,7 @@ private void runPredict(
return;
} catch (Exception e) {
log.error("Failed to predict model " + modelId, e);
handlePredictFailure(mlTask, internalListener, e, false, modelId, actionName);
handlePredictFailure(mlTask, internalListener, e, shouldTrackRemoteFailure(e), modelId, actionName);
return;
}
} else if (FunctionName.needDeployFirst(algorithm)) {
Expand Down Expand Up @@ -604,4 +604,22 @@ public void validateOutputSchema(String modelId, ModelTensorOutput output) {
}
}
}

boolean shouldTrackRemoteFailure(Exception e) {
// Don't track failures for user configuration issues
if (e instanceof IllegalArgumentException) {
return false;
}

// Don't track any 4xx client errors (user/configuration issues)
if (e instanceof OpenSearchStatusException) {
RestStatus status = ((OpenSearchStatusException) e).status();
if (status.getStatus() >= 400 && status.getStatus() < 500) {
return false;
}
}

// Track failures for infrastructure/service issues
return true;
}
}
22 changes: 22 additions & 0 deletions plugin/src/test/java/org/opensearch/ml/task/MLPredictTaskRunnerTests.java
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -50,6 +50,7 @@
import org.opensearch.core.action.ActionListener;
import org.opensearch.core.common.bytes.BytesReference;
import org.opensearch.core.common.transport.TransportAddress;
import org.opensearch.core.rest.RestStatus;
import org.opensearch.core.xcontent.ToXContent;
import org.opensearch.core.xcontent.XContentBuilder;
import org.opensearch.index.get.GetResult;
Expand Down Expand Up @@ -721,4 +722,25 @@ private void setupMocks(boolean runOnLocalNode, boolean failedToParseQueryInput,
}).when(client).get(any(), any());
}
}

public void testShouldTrackRemoteFailure() {
// Test IllegalArgumentException - should not track
assertFalse(taskRunner.shouldTrackRemoteFailure(new IllegalArgumentException("Invalid argument")));

// Test OpenSearchStatusException with 4xx status codes - should not track
assertFalse(taskRunner.shouldTrackRemoteFailure(new OpenSearchStatusException("Bad request", RestStatus.BAD_REQUEST)));
assertFalse(taskRunner.shouldTrackRemoteFailure(new OpenSearchStatusException("Unauthorized", RestStatus.UNAUTHORIZED)));
assertFalse(taskRunner.shouldTrackRemoteFailure(new OpenSearchStatusException("Forbidden", RestStatus.FORBIDDEN)));
assertFalse(taskRunner.shouldTrackRemoteFailure(new OpenSearchStatusException("Not found", RestStatus.NOT_FOUND)));

// Test OpenSearchStatusException with 5xx status codes - should track
assertTrue(taskRunner.shouldTrackRemoteFailure(new OpenSearchStatusException("Server error", RestStatus.INTERNAL_SERVER_ERROR)));
assertTrue(
taskRunner.shouldTrackRemoteFailure(new OpenSearchStatusException("Service unavailable", RestStatus.SERVICE_UNAVAILABLE))
);

// Test other exceptions - should track
assertTrue(taskRunner.shouldTrackRemoteFailure(new RuntimeException("Runtime error")));
assertTrue(taskRunner.shouldTrackRemoteFailure(new IOException("IO error")));
}
}
Loading