fix(test-case): update 5000 tables test case configuration #9843
Conversation
This file contains 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
vponomaryov
added
backport/2024.2
roydahan
reviewed
Jan 16, 2025
roydahan
previously approved these changes
Jan 16, 2025
List of changes: - Disable per-table metrics due to significant perf impact. - Enable cluster health checks which work with this case just fine. - Decrease the nemesis interval from 60 minutes to just 3 keeping in mind that health checks will take some time too. - Reduce stress time for each of the 5000 commands. Having 20 minutes per cmd we will get about 1.5 days long test runs instead of the 2.5 days. - Reduce number of loaders from 5 to 3 to use resources more efficiently. In current case the bottleneck is the RAM. Note that this scenario hits following bug: - scylladb/scylla-enterprise#5093 If 'destroy_data_then_repair' nemesis gets triggered aganst the setup of this scenario.
vponomaryov
force-pushed
the
update-5000-tables-senario
branch
from
878caa3 to
6db9fa9
Compare
|
Is this with tablets or with vnodes? (we probably need for both and compare between them) |
This was referenced Jan 19, 2025
test can run with both, and the issue mentioned above happened with both you can see the summerie here: but it doesn't seem like anyone is attending to the issue raised by @vponomaryov |
scylladbbot
added
backport/6.1-done
vponomaryov
added a commit
to vponomaryov/scylla-cluster-tests
that referenced
this pull request
Jan 21, 2025
The 'test_longevity.py::test_test_user_batch_custom_time' unit test uses the 'test-cases/scale/longevity-5000-tables.yaml' config file for running a short longevity test which triggers a nemesis. If health checks are enabled then the "nemesis call" runs much longer while health checks are completed 2 times - before and after the nemesis. And while is it ongoing the nemesis lock gets held. And the problem with it is that it runs even after finish of this test. So, any another unit test which tries to get a nemesis lock will stumble upon a held lock for 10+ minutes. It started happening after the merge a PR [1] which enabled health checks in the mentioned config file. The affected test is following: - test_nemesis.py::test_list_nemesis_of_added_disrupt_methods Alphabetically it runs after the 'test_longevity.py::test_test_user_batch_custom_time' one. So, disable health checks to avoid side-effects and doing redundant stuff which was not planned when the test was written. [1] scylladb#9843
vponomaryov
added a commit
to vponomaryov/scylla-cluster-tests
that referenced
this pull request
Jan 21, 2025
The 'test_longevity.py::test_test_user_batch_custom_time' unit test uses the 'test-cases/scale/longevity-5000-tables.yaml' config file for running a short longevity test which triggers a nemesis. If health checks are enabled then the "nemesis call" runs much longer while health checks are completed 2 times - before and after the nemesis. And while it is ongoing the nemesis lock gets held. And the problem with it is that it runs even after finish of this test. So, any another unit test which tries to get a nemesis lock will stumble upon a held lock for 10+ minutes. It started happening after the merge a PR [1] which enabled health checks in the mentioned config file. The affected test is following: - test_nemesis.py::test_list_nemesis_of_added_disrupt_methods Alphabetically it runs after the 'test_longevity.py::test_test_user_batch_custom_time' one. So, disable health checks to avoid side-effects and doing redundant stuff which was not planned when the test was written. [1] scylladb#9843
vponomaryov
added a commit
to vponomaryov/scylla-cluster-tests
that referenced
this pull request
Jan 21, 2025
The 'test_longevity.py::test_test_user_batch_custom_time' unit test uses the 'test-cases/scale/longevity-5000-tables.yaml' config file for running a short longevity test which triggers a nemesis. If health checks are enabled then the "nemesis call" runs much longer while health checks are completed 2 times - before and after the nemesis. And while it is ongoing the nemesis lock gets held. And the problem with it is that it runs even after finish of this test. So, any another unit test which tries to get a nemesis lock will stumble upon a held lock for 10+ minutes. It started happening after the merge of the PR [1] which enabled health checks in the mentioned config file. The affected test is following: - test_nemesis.py::test_list_nemesis_of_added_disrupt_methods Alphabetically it runs after the 'test_longevity.py::test_test_user_batch_custom_time' one. So, disable health checks to avoid side-effects and doing redundant stuff which was not planned when the test was written. [1] scylladb#9843
vponomaryov
added a commit
to vponomaryov/scylla-cluster-tests
that referenced
this pull request
Jan 21, 2025
The 'test_longevity.py::test_test_user_batch_custom_time' unit test uses the 'test-cases/scale/longevity-5000-tables.yaml' config file for running a short longevity test which triggers a nemesis. If health checks are enabled then the "nemesis call" runs much longer while health checks are completed 2 times - before and after the nemesis. And while it is ongoing the nemesis lock gets held. And the problem with it is that it runs even after finish of this test. So, any another unit test which tries to get a nemesis lock will stumble upon a held lock for 10+ minutes. It started happening after the merge of the PR [1] which enabled health checks in the mentioned config file. The affected test is following: - test_nemesis.py::test_list_nemesis_of_added_disrupt_methods Alphabetically it runs after the 'test_longevity.py::test_test_user_batch_custom_time' one. So, disable health checks to avoid side-effects and doing redundant stuff which were not planned when the test was written. [1] scylladb#9843
2 tasks
vponomaryov
added a commit
to vponomaryov/scylla-cluster-tests
that referenced
this pull request
Jan 21, 2025
The 'test_longevity.py::test_test_user_batch_custom_time' unit test uses the 'test-cases/scale/longevity-5000-tables.yaml' config file for running a short longevity test which triggers a nemesis. If health checks are enabled then the "nemesis call" runs much longer while health checks are completed 2 times - before and after the nemesis. And while it is ongoing the nemesis lock gets held. And the problem with it is that it runs even after finish of this test. So, any another unit test which tries to get a nemesis lock will stumble upon a held lock for 10+ minutes. It started happening after the merge of the PR [1] which enabled health checks in the mentioned config file. The affected test is following: - test_nemesis.py::test_list_nemesis_of_added_disrupt_methods Alphabetically it runs after the 'test_longevity.py::test_test_user_batch_custom_time' one. So, disable health checks to avoid side-effects and doing redundant stuff which were not planned when the test was written. [1] scylladb#9843
fruch
pushed a commit
that referenced
this pull request
Jan 21, 2025
The 'test_longevity.py::test_test_user_batch_custom_time' unit test uses the 'test-cases/scale/longevity-5000-tables.yaml' config file for running a short longevity test which triggers a nemesis. If health checks are enabled then the "nemesis call" runs much longer while health checks are completed 2 times - before and after the nemesis. And while it is ongoing the nemesis lock gets held. And the problem with it is that it runs even after finish of this test. So, any another unit test which tries to get a nemesis lock will stumble upon a held lock for 10+ minutes. It started happening after the merge of the PR [1] which enabled health checks in the mentioned config file. The affected test is following: - test_nemesis.py::test_list_nemesis_of_added_disrupt_methods Alphabetically it runs after the 'test_longevity.py::test_test_user_batch_custom_time' one. So, disable health checks to avoid side-effects and doing redundant stuff which were not planned when the test was written. [1] #9843
scylladbbot
pushed a commit
to scylladbbot/scylla-cluster-tests
that referenced
this pull request
Jan 21, 2025
The 'test_longevity.py::test_test_user_batch_custom_time' unit test uses the 'test-cases/scale/longevity-5000-tables.yaml' config file for running a short longevity test which triggers a nemesis. If health checks are enabled then the "nemesis call" runs much longer while health checks are completed 2 times - before and after the nemesis. And while it is ongoing the nemesis lock gets held. And the problem with it is that it runs even after finish of this test. So, any another unit test which tries to get a nemesis lock will stumble upon a held lock for 10+ minutes. It started happening after the merge of the PR [1] which enabled health checks in the mentioned config file. The affected test is following: - test_nemesis.py::test_list_nemesis_of_added_disrupt_methods Alphabetically it runs after the 'test_longevity.py::test_test_user_batch_custom_time' one. So, disable health checks to avoid side-effects and doing redundant stuff which were not planned when the test was written. [1] scylladb#9843 (cherry picked from commit a78d65f)
Closed
2 tasks
scylladbbot
pushed a commit
to scylladbbot/scylla-cluster-tests
that referenced
this pull request
Jan 21, 2025
The 'test_longevity.py::test_test_user_batch_custom_time' unit test uses the 'test-cases/scale/longevity-5000-tables.yaml' config file for running a short longevity test which triggers a nemesis. If health checks are enabled then the "nemesis call" runs much longer while health checks are completed 2 times - before and after the nemesis. And while it is ongoing the nemesis lock gets held. And the problem with it is that it runs even after finish of this test. So, any another unit test which tries to get a nemesis lock will stumble upon a held lock for 10+ minutes. It started happening after the merge of the PR [1] which enabled health checks in the mentioned config file. The affected test is following: - test_nemesis.py::test_list_nemesis_of_added_disrupt_methods Alphabetically it runs after the 'test_longevity.py::test_test_user_batch_custom_time' one. So, disable health checks to avoid side-effects and doing redundant stuff which were not planned when the test was written. [1] scylladb#9843 (cherry picked from commit a78d65f)
Merged
2 tasks
scylladbbot
pushed a commit
to scylladbbot/scylla-cluster-tests
that referenced
this pull request
Jan 21, 2025
The 'test_longevity.py::test_test_user_batch_custom_time' unit test uses the 'test-cases/scale/longevity-5000-tables.yaml' config file for running a short longevity test which triggers a nemesis. If health checks are enabled then the "nemesis call" runs much longer while health checks are completed 2 times - before and after the nemesis. And while it is ongoing the nemesis lock gets held. And the problem with it is that it runs even after finish of this test. So, any another unit test which tries to get a nemesis lock will stumble upon a held lock for 10+ minutes. It started happening after the merge of the PR [1] which enabled health checks in the mentioned config file. The affected test is following: - test_nemesis.py::test_list_nemesis_of_added_disrupt_methods Alphabetically it runs after the 'test_longevity.py::test_test_user_batch_custom_time' one. So, disable health checks to avoid side-effects and doing redundant stuff which were not planned when the test was written. [1] scylladb#9843 (cherry picked from commit a78d65f)
2 tasks
scylladbbot
pushed a commit
to scylladbbot/scylla-cluster-tests
that referenced
this pull request
Jan 21, 2025
The 'test_longevity.py::test_test_user_batch_custom_time' unit test uses the 'test-cases/scale/longevity-5000-tables.yaml' config file for running a short longevity test which triggers a nemesis. If health checks are enabled then the "nemesis call" runs much longer while health checks are completed 2 times - before and after the nemesis. And while it is ongoing the nemesis lock gets held. And the problem with it is that it runs even after finish of this test. So, any another unit test which tries to get a nemesis lock will stumble upon a held lock for 10+ minutes. It started happening after the merge of the PR [1] which enabled health checks in the mentioned config file. The affected test is following: - test_nemesis.py::test_list_nemesis_of_added_disrupt_methods Alphabetically it runs after the 'test_longevity.py::test_test_user_batch_custom_time' one. So, disable health checks to avoid side-effects and doing redundant stuff which were not planned when the test was written. [1] scylladb#9843 (cherry picked from commit a78d65f)
2 tasks
fruch
pushed a commit
that referenced
this pull request
Jan 21, 2025
The 'test_longevity.py::test_test_user_batch_custom_time' unit test uses the 'test-cases/scale/longevity-5000-tables.yaml' config file for running a short longevity test which triggers a nemesis. If health checks are enabled then the "nemesis call" runs much longer while health checks are completed 2 times - before and after the nemesis. And while it is ongoing the nemesis lock gets held. And the problem with it is that it runs even after finish of this test. So, any another unit test which tries to get a nemesis lock will stumble upon a held lock for 10+ minutes. It started happening after the merge of the PR [1] which enabled health checks in the mentioned config file. The affected test is following: - test_nemesis.py::test_list_nemesis_of_added_disrupt_methods Alphabetically it runs after the 'test_longevity.py::test_test_user_batch_custom_time' one. So, disable health checks to avoid side-effects and doing redundant stuff which were not planned when the test was written. [1] #9843 (cherry picked from commit a78d65f)
fruch
pushed a commit
that referenced
this pull request
Jan 21, 2025
The 'test_longevity.py::test_test_user_batch_custom_time' unit test uses the 'test-cases/scale/longevity-5000-tables.yaml' config file for running a short longevity test which triggers a nemesis. If health checks are enabled then the "nemesis call" runs much longer while health checks are completed 2 times - before and after the nemesis. And while it is ongoing the nemesis lock gets held. And the problem with it is that it runs even after finish of this test. So, any another unit test which tries to get a nemesis lock will stumble upon a held lock for 10+ minutes. It started happening after the merge of the PR [1] which enabled health checks in the mentioned config file. The affected test is following: - test_nemesis.py::test_list_nemesis_of_added_disrupt_methods Alphabetically it runs after the 'test_longevity.py::test_test_user_batch_custom_time' one. So, disable health checks to avoid side-effects and doing redundant stuff which were not planned when the test was written. [1] #9843 (cherry picked from commit a78d65f)
scylladbbot
added
backport/2024.2-done
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
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.
List of changes:
Note that this scenario hits following bug:
If
destroy_data_then_repairnemesis gets triggered aganst the setup of this scenario.Testing
PR pre-checks (self review)
backportlabelsReminders
sdcm/sct_config.py)unit-test/folder)