[SERVER-57273] remove the flow_control_replica_set.js test Created: 27/May/21  Updated: 29/Oct/23  Resolved: 10/Jun/21

Status: Closed
Project: Core Server
Component/s: Testing Infrastructure
Affects Version/s: None
Fix Version/s: 4.2.15, 4.4.7, 5.0.0-rc3, 5.1.0-rc0

Type: Improvement Priority: Major - P3
Reporter: Benety Goh Assignee: Benety Goh
Resolution: Fixed Votes: 0
Labels: None
Remaining Estimate: Not Specified
Time Spent: Not Specified
Original Estimate: Not Specified

Issue Links:
Backports
Depends
Related
is related to SERVER-40447 Add Flow Control mechanism unit/integ... Closed
is related to SERVER-41622 Reduce noisiness in flow_control_repl... Closed
Backwards Compatibility: Fully Compatible
Backport Requested:
v5.0, v4.4, v4.2
Sprint: Execution Team 2021-06-28
Participants:
Linked BF Score: 18

 Description   

We can reduce the number of failures in this test due to machine flakiness by not running it on certain platforms, such as Amazon Linux and macOS. This can be done by adding the following tags:

  • incompatible_with_amazon_linux
  • incompatible_with_macos


 Comments   
Comment by Vivian Ge (Inactive) [ 06/Oct/21 ]

Updating the fixversion since branching activities occurred yesterday. This ticket will be in rc0 when it’s been triggered. For more active release information, please keep an eye on #server-release. Thank you!

Comment by Githook User [ 30/Jun/21 ]

Author:

{'name': 'Benety Goh', 'email': 'benety@mongodb.com', 'username': 'benety'}

Message: SERVER-57273 remove flow_control_replica_set.js

This test fails sporadically due to causes unrelated to the feature this
test was originally intended to test. Removing this test for now.

For C++ unit test coverage, refer to mongo/db/storage/flow_control_test.cpp.

(cherry picked from commit 17691de3ce474d8d9c476d1ee99590d8c3edd291)
Branch: v4.2
https://github.com/mongodb/mongo/commit/f35d9bd6beb8259fcefbe6e7dfb98c745a37f285

Comment by Githook User [ 29/Jun/21 ]

Author:

{'name': 'Benety Goh', 'email': 'benety@mongodb.com', 'username': 'benety'}

Message: SERVER-57273 remove flow_control_replica_set.js

This test fails sporadically due to causes unrelated to the feature this
test was originally intended to test. Removing this test for now.

For C++ unit test coverage, refer to mongo/db/storage/flow_control_test.cpp.

(cherry picked from commit 17691de3ce474d8d9c476d1ee99590d8c3edd291)
Branch: v4.4
https://github.com/mongodb/mongo/commit/22139444ca9372526900fcd454345f8c8a7f2988

Comment by Githook User [ 18/Jun/21 ]

Author:

{'name': 'Benety Goh', 'email': 'benety@mongodb.com', 'username': 'benety'}

Message: SERVER-57273 remove flow_control_replica_set.js

This test fails sporadically due to causes unrelated to the feature this
test was originally intended to test. Removing this test for now.

For C++ unit test coverage, refer to mongo/db/storage/flow_control_test.cpp.

(cherry picked from commit 17691de3ce474d8d9c476d1ee99590d8c3edd291)
Branch: v5.0
https://github.com/mongodb/mongo/commit/750aad5c3c0c287fed7fbda7b704ac2162614bf7

Comment by Githook User [ 10/Jun/21 ]

Author:

{'name': 'Benety Goh', 'email': 'benety@mongodb.com', 'username': 'benety'}

Message: SERVER-57273 remove flow_control_replica_set.js

This test fails sporadically due to causes unrelated to the feature this
test was originally intended to test. Removing this test for now.

For C++ unit test coverage, refer to mongo/db/storage/flow_control_test.cpp.
Branch: master
https://github.com/mongodb/mongo/commit/17691de3ce474d8d9c476d1ee99590d8c3edd291

Comment by Benety Goh [ 08/Jun/21 ]

Here is a snippet of the test on the current 4.2 HEAD. I'm sure if we run the test long enough, we can hope to smooth out the results. But at that point, we may as well rewrite it as a system performance workload (which I hope there's already one present for the feature).

result = benchRun({
    host: primary.host,
    seconds: 5,
    parallel: 5,
    ops: [{op: "insert", ns: "foo.bar", doc: {field: "value"}}]
});
jsTestLog({ThrottledRun: result, ThrottedRate: throttledRate});
let maxAllowedRate = 1.5 * throttledRate;
let minAllowedRate = 0.5 * throttledRate;
assert.gt(result["insert"], minAllowedRate);
assert.lt(result["insert"], maxAllowedRate);

Comment by Andrew Morrow (Inactive) [ 02/Jun/21 ]

That'd be more convincing to me if it was just macOS, which is known to be rather slow, but looking at the build failure, it appears to have also happened on RHEL 7 and AL2 and Ubuntu 18.04. And I don't see any reason that plain AL should be any slower than any other linux distro. I think we should consider whether there may be other ways to stabilize the test, rather than trying to paper over it with instance based suppressions.

Comment by Benety Goh [ 01/Jun/21 ]

acm, from what I can observe in the logs and my understanding of what this test does, the test is essentially a performance test. There's a certain amount of heuristics incorporated in the test to determine what is acceptable for a successful test run.

 

 

The linked tickets point to past efforts to make this test more tolerant of machine environments that may not match those of the required builders.

Comment by Andrew Morrow (Inactive) [ 01/Jun/21 ]

benety.goh - Can you elaborate on what sort of machine flakiness we are observing with this test? I'd rather fix the machine flakiness than disable tests.

Generated at Thu Feb 08 05:41:25 UTC 2024 using Jira 9.7.1#970001-sha1:2222b88b221c4928ef0de3161136cc90c8356a66.