etcd/tests/robustness

etcd Robustness Testing

Purpose of etcd robustness tests is to validate that etcd upholds KV API guarantees and watch API guarantees under any condition or failure.

Robustness tests achieve that comparing etcd cluster behavior against a simplified model. Multiple test encompass different etcd cluster setups, client traffic types and failures experienced by cluster. During a single test we create a cluster and inject failures while sending and recording client traffic. Correctness is validated by running collected history of client operations against the etcd model and a set of validators. Upon failure tests generate a report that can be used to attribute whether failure was caused by bug in etcd or test framework.

Running locally

1. Build etcd with failpoints

   make gofail-enable
   make build
   make gofail-disable
   

2. Run the tests

   make test-robustness
   

   Optionally you can pass environment variables:

   • GO_TEST_FLAGS - to pass additional arguments to go test. It is recommended to run tests multiple times with failfast enabled. this can be done by setting GO_TEST_FLAGS='--count=100 --failfast'.
   • EXPECT_DEBUG=true - to get logs from the cluster.
   • RESULTS_DIR - to change location where results report will be saved.
   • PERSIST_RESULTS - to persist the results report of the test. By default this will not be persisted in the case of a successful run.

Re-evaluate existing report

Robustness test validation is constantly changing and improving. Errors in etcd model could be causing false positives, which makes the ability to re-evaluate the reports after we fix the issue important.

Note: Robustness test report format is not stable, and it's expected that not all old reports can be re-evaluated using the newest version.

1. Identify location of the robustness test report.

   Note: By default robustness test report is only generated for failed test.

   • For local runs: this would be by identifying log line, in the following example that would be /tmp/TestRobustnessExploratory_Etcd_HighTraffic_ClusterOfSize1:

     logger.go:146: 2024-04-08T09:45:27.734+0200 INFO    Saving robustness test report   {"path": "/tmp/TestRobustnessExploratory_Etcd_HighTraffic_ClusterOfSize1"}
     

   • For remote runs on CI: you need to go to the Github Actions page, download one of the Artifacts, extract it locally.

     

     Each directory will be prefixed by TestRobustness each containing a robustness test report.

     

     Pick one of the directories within the archive corresponding to the failed test scenario. The largest directory by size usually corresponds to the failed scenario. If you are not sure, you may check which scenario failed in the test logs.

2. Copy the robustness report directory into the testdata directory.

   The testdata directory can contain multiple robustness test reports. The name of the report directory doesn't matter, as long as it's unique to prevent clashing with reports already present in testdata directory. For example path for history.html file could look like $REPO_ROOT/tests/robustness/testdata/v3.5_failure_24_April/history.html.

3. Run make test-robustness-reports to validate all reports in the testdata directory.

Analysing failure

If robustness tests fails we want to analyse the report to confirm if the issue is on etcd side. Location of this report is included in test logs. One of log lines should look like:

    history.go:34: Model is not linearizable
    logger.go:130: 2023-03-18T12:18:03.244+0100 INFO    Saving member data dir  {"member": "TestRobustnessIssue14370-test-0", "path": "/tmp/TestRobustness_Issue14370/TestRobustnessIssue14370-test-0"}
    logger.go:130: 2023-03-18T12:18:03.244+0100 INFO    Saving watch responses  {"path": "/tmp/TestRobustness_Issue14370/TestRobustnessIssue14370-test-0/responses.json"}
    logger.go:130: 2023-03-18T12:18:03.247+0100 INFO    Saving watch events     {"path": "/tmp/TestRobustness_Issue14370/TestRobustnessIssue14370-test-0/events.json"}
    logger.go:130: 2023-05-15T17:42:37.792+0200 INFO    Saving operation history        {"path": "/tmp/TestRobustness_ClusterOfSize3_Kubernetes/client-1/operations.json"}
    logger.go:130: 2023-05-15T17:42:37.793+0200 INFO    Saving watch responses  {"path": "/tmp/TestRobustness_ClusterOfSize3_Kubernetes/client-2/watch.json"}
    logger.go:130: 2023-03-18T12:18:03.256+0100 INFO    Saving visualization    {"path": "/tmp/TestRobustness_Issue14370/history.html"}

Report includes multiple types of files:

• Member db files, can be used to verify disk/memory corruption.
• Watch responses saved as json, can be used to validate watch API guarantees.
• Operation history saved as both html visualization and a json, can be used to validate KV API guarantees.

Example analysis of linearization issue

Let's reproduce and analyse robustness test report for issue #14370. To reproduce the issue by yourself run make test-robustness-issue14370. After a couple of tries robustness tests should fail with a log Model is not linearizable and save report locally.

Lineralization issues are easiest to analyse via history visualization. Open /tmp/TestRobustness_Issue14370/history.html file in your browser. Jump to the error in linearization by clicking [ jump to first error ] on the top of the page.

You should see a graph similar to the one on the image below.

Last correct request (connected with grey line) is a Put request that succeeded and got revision 168. All following requests are invalid (connected with red line) as they have revision 167. Etcd guarantee that revision is non-decreasing, so this shows a bug in etcd as there is no way revision should decrease. This is consistent with the root cause of #14370 as it was issue with process crash causing last write to be lost.

Example analysis of watch issue

Let's reproduce and analyse robustness test report for issue #15271. To reproduce the issue by yourself run make test-robustness-issue15271. After a couple of tries robustness tests should fail with a logs Broke watch guarantee and save report locally.

Watch issues are easiest to analyse by reading the recorded watch history. Watch history is recorded for each client separated in different subdirectory under /tmp/TestRobustness_Issue15271/ Open watch.json for client mentioned in log Broke watch guarantee. For example for client 14 open /tmp/TestRobustness_Issue15271/client-14/watch.json.

Each line consists of json blob corresponding to single watch response observed by client. Look for lines with mod_revision equal to revision mentioned in the first log with Broke watch guarantee You should see two lines where the mod_revision decreases like ones below:

{"Events":[{"Op":{"Type":"put","Key":"5","WithPrefix":false,"Limit":0,"Value":{"Value":"357","Hash":0},"LeaseID":0},"Revision":335}],"IsProgressNotify":false,"Revision":335,"Time":1050415777}
{"Events":[{"Op":{"Type":"put","Key":"1","WithPrefix":false,"Limit":0,"Value":{"Value":"24","Hash":0},"LeaseID":0},"Revision":24}, ...

Up to the first line the revision of events within responses only increased up to a value of 335. However, the following line includes an event with revision equal 24. If you follow the revision throughout the file you should notice that watch replayed revisions second time. This is incorrect and breaks Ordered and Unique watch API guarantees. This is consistent with the root cause of #14370 where member reconnecting to cluster will incorrectly resend revisions.