You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
Currently many of our integration tests for our processors (text embedding, text image embedding, sparse encoding) follow the following pattern to create and cleanup resources
public void processor_integrationTest() {
try {
// load model
// create pipeline/index
// do tests/asserts
} finally {
// cleanup pipeline, index, model
}
}
However, this causes issues like #1091 where if the integration tests fail in the try block before creating the pipeline, it will enter the finally block and attempt to delete the nonexistent pipeline. This causes an exception in the cleanup and masks the true exception, such as #1093 where the root issue is a memory circuit breaker trigger which is unclear from the test report.
We should rework our integration test cleanup so we can:
Log errors more clearly during resource cleanup to avoid confusion
Resource cleanup after every test rather than a part of every test, so cleanup won't directly cause test failures
The text was updated successfully, but these errors were encountered:
Currently many of our integration tests for our processors (text embedding, text image embedding, sparse encoding) follow the following pattern to create and cleanup resources
However, this causes issues like #1091 where if the integration tests fail in the try block before creating the pipeline, it will enter the finally block and attempt to delete the nonexistent pipeline. This causes an exception in the cleanup and masks the true exception, such as #1093 where the root issue is a memory circuit breaker trigger which is unclear from the test report.
We should rework our integration test cleanup so we can:
The text was updated successfully, but these errors were encountered: