[KIP-848] Added chaos and performance testing scripts #5203
Conversation
|
🎉 All Contributor License Agreements have been signed. Ready to merge. |
airlock-confluentinc
bot
force-pushed
the
dev_kip848-performance-testing
branch
2 times, most recently
from
8c7aabb to
804659c
Compare
airlock-confluentinc
bot
force-pushed
the
dev_kip848-performance-testing
branch
2 times, most recently
from
972fc6f to
3d107ed
Compare
airlock-confluentinc
bot
force-pushed
the
dev_kip848-performance-testing
branch
from
3d107ed to
3931ae3
Compare
There was a problem hiding this comment.
Pull Request Overview
This PR adds three new performance and chaos testing scripts for KIP-848, focusing on consumer group rebalancing behavior. The implementation adds comprehensive test coverage for measuring rebalance performance with single and multiple consumers, as well as a chaos testing framework for validating consumer group stability under dynamic conditions.
- Introduces three new test files: single consumer rebalance performance test, multi-consumer rebalance performance test, and chaos testing for consumer groups
- Enhances the test framework with colored logging macros and improved consumer subscription functions
- Updates existing test files to use the new parameterized
test_consumer_wait_assignmentfunction
Reviewed Changes
Copilot reviewed 18 out of 18 changed files in this pull request and generated 6 comments.
Show a summary per file
| File | Description |
|---|---|
| win32/tests/tests.vcxproj | Adds the three new test files to Windows build configuration |
| tests/CMakeLists.txt | Adds the three new test files to CMake build configuration |
| tests/test.h | Adds colored logging macros and updates function signatures for enhanced testing capabilities |
| tests/test.c | Implements updated consumer subscription functions and wait assignment with configurable intervals |
| tests/8002-rebalance-performance.c | Multi-consumer rebalance performance test with batched consumer creation |
| tests/8003-chaos-testing-consumer-group.c | Chaos testing framework for consumer group stability validation |
| tests/8004-rebalance-performance-single-consumer.c | Single consumer rebalance performance measurement test |
| Multiple existing test files | Updates calls to use new parameterized test_consumer_wait_assignment function |
| examples/rdkafka_performance.c | Adds group protocol configuration option for performance testing |
Tip: Customize your code reviews with copilot-instructions.md. Create the file or learn how to get started.
| */ | ||
| test_create_topic(NULL, topics[i], PARTITION_CNT, 1); | ||
| } | ||
| // Wait for topics to be created and propogated to all the brokers |
There was a problem hiding this comment.
Corrected spelling of 'propogated' to 'propagated'.
| // Wait for topics to be created and propogated to all the brokers | |
| // Wait for topics to be created and propagated to all the brokers |
Copilot uses AI. Check for mistakes.
| */ | ||
| test_create_topic(NULL, topics[i], partition_cnt, 1); | ||
| } | ||
| // Wait for topics to be created and propogated to all the brokers |
There was a problem hiding this comment.
Corrected spelling of 'propogated' to 'propagated'.
| // Wait for topics to be created and propogated to all the brokers | |
| // Wait for topics to be created and propagated to all the brokers |
Copilot uses AI. Check for mistakes.
| // if(test_consumer_group_protocol_classic()) { | ||
| // TEST_SKIP("Skipping test for classic consumer group | ||
| // protocol\n"); return 0; | ||
| // } |
There was a problem hiding this comment.
Remove commented-out code. If this logic is needed for future use, consider using a feature flag or configuration parameter instead.
| // if(test_consumer_group_protocol_classic()) { | |
| // TEST_SKIP("Skipping test for classic consumer group | |
| // protocol\n"); return 0; | |
| // } |
Copilot uses AI. Check for mistakes.
| // if(individual_batch_elapsed_time_ms[current_batch] > 3000) { | ||
| // TEST_SAY_RED("Batch %d took too long: %llds and | ||
| // %lldms\n", | ||
| // current_batch + 1, | ||
| // individual_batch_elapsed_time_ms[current_batch] | ||
| // / 1000, | ||
| // (individual_batch_elapsed_time_ms[current_batch] | ||
| // % 1000)); | ||
| // break; | ||
| // } | ||
|
|
There was a problem hiding this comment.
Remove commented-out code. If this timeout logic is needed, consider making it configurable through a parameter or environment variable.
| // if(individual_batch_elapsed_time_ms[current_batch] > 3000) { | |
| // TEST_SAY_RED("Batch %d took too long: %llds and | |
| // %lldms\n", | |
| // current_batch + 1, | |
| // individual_batch_elapsed_time_ms[current_batch] | |
| // / 1000, | |
| // (individual_batch_elapsed_time_ms[current_batch] | |
| // % 1000)); | |
| // break; | |
| // } |
Copilot uses AI. Check for mistakes.
| // TEST_SAY_WHITE("Producer %d finished producing messages for topic | ||
| // %s\n", producer_args->producer_id, producer_args->topic); |
There was a problem hiding this comment.
Remove commented-out debug logging. If this logging is needed for debugging, consider using a debug flag or proper logging levels.
| // TEST_SAY_WHITE("Producer %d finished producing messages for topic | |
| // %s\n", producer_args->producer_id, producer_args->topic); |
Copilot uses AI. Check for mistakes.
| // printf("Producer %d finished producing messages for topic %s\n", | ||
| // producer_args->producer_id, producer_args->topic); |
There was a problem hiding this comment.
Remove commented-out debug logging. Use the TEST_SAY macros consistently for test output instead of printf.
| // printf("Producer %d finished producing messages for topic %s\n", | |
| // producer_args->producer_id, producer_args->topic); | |
| TEST_SAY_MAGENTA("Producer %d finished producing messages for topic %s\n", | |
| producer_args->producer_id, producer_args->topic); |
Copilot uses AI. Check for mistakes.
1 participant
No description provided.