Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Final A/B-Tests #4139

Merged
merged 9 commits into from
Oct 2, 2023
Merged

Final A/B-Tests #4139

merged 9 commits into from
Oct 2, 2023

Conversation

roypat
Copy link
Contributor

@roypat roypat commented Sep 28, 2023

Changes

Adds A/B compatible versions of the vsock and tcp throughput tests. Also includes the following small refractorings/improvements:

  • Declare specific tests are particularly unstable. These are the tests that already have significant delta values associated with them in the current testing setup.
  • Correctly associate p-values with metrics when submitting to Cloudwatch
  • Correctly propagate CW metadata when reemitting captured logs (namespace and log stream name)
  • Only run A/B tests if rust/python code was modified.
  • Merge the two network tests into one file, because they are significantly simpler now

License Acceptance

By submitting this pull request, I confirm that my contribution is made under
the terms of the Apache 2.0 license. For more information on following
Developer Certificate of Origin and signing off your commits, please check
CONTRIBUTING.md.

PR Checklist

  • If a specific issue led to this PR, this PR closes the issue.
  • The description of changes is clear and encompassing.
  • Any required documentation changes (code and docs) are included in this PR.
  • API changes follow the Runbook for Firecracker API changes.
  • User-facing changes are mentioned in CHANGELOG.md.
  • All added/changed functionality is tested.
  • New TODOs link to an issue.
  • Commits meet contribution quality standards.

  • This functionality cannot be added in rust-vmm.

@codecov
Copy link

codecov bot commented Sep 28, 2023

Codecov Report

All modified lines are covered by tests ✅

Comparison is base (a376ab1) 83.10% compared to head (85dae0f) 83.10%.

Additional details and impacted files
@@           Coverage Diff           @@
##             main    #4139   +/-   ##
=======================================
  Coverage   83.10%   83.10%           
=======================================
  Files         225      225           
  Lines       28604    28604           
=======================================
  Hits        23771    23771           
  Misses       4833     4833           
Flag Coverage Δ
4.14-c7g.metal 78.67% <ø> (ø)
4.14-m5d.metal 80.47% <ø> (ø)
4.14-m6a.metal 79.61% <ø> (ø)
4.14-m6g.metal 78.67% <ø> (ø)
4.14-m6i.metal 80.45% <ø> (ø)
5.10-c7g.metal 81.57% <ø> (ø)
5.10-m5d.metal 83.13% <ø> (ø)
5.10-m6a.metal 82.37% <ø> (ø)
5.10-m6g.metal 81.57% <ø> (ø)
5.10-m6i.metal 83.11% <ø> (ø)
6.1-c7g.metal 81.57% <ø> (+<0.01%) ⬆️
6.1-m5d.metal 83.13% <ø> (ø)
6.1-m6a.metal 82.37% <ø> (ø)
6.1-m6g.metal 81.57% <ø> (ø)
6.1-m6i.metal 83.11% <ø> (ø)

Flags with carried forward coverage won't be shown. Click here to find out more.

☔ View full report in Codecov by Sentry.
📢 Have feedback on the report? Share it here.

@roypat roypat force-pushed the most-ab branch 4 times, most recently from 138d51c to faa15d2 Compare September 29, 2023 09:07
@roypat roypat marked this pull request as ready for review September 29, 2023 09:07
@roypat roypat force-pushed the most-ab branch 3 times, most recently from b809012 to 4e6ec9f Compare September 29, 2023 09:28
zulinx86
zulinx86 previously approved these changes Sep 29, 2023
tests/integration_tests/performance/test_vsock_ab.py Outdated Show resolved Hide resolved
tests/integration_tests/performance/test_vsock_ab.py Outdated Show resolved Hide resolved
Similar to the snapshot, network latency and block tests, our vsock
throughput test adapted to work with A/B-testing.

Signed-off-by: Patrick Roy <[email protected]>
Converts the existing network tcp throughput test into an A/B-compatible
TCP throughput test. This test is added to test_network_ab.py, since
the converted test is short enough that we can collect all the network
related tests in a single file.

Signed-off-by: Patrick Roy <[email protected]>
Without this information, it is impossible to tell apart p-values for
different metrics emitted from the same test.

Signed-off-by: Patrick Roy <[email protected]>
Otherwise, the reemitted metrics will carry a namespace of "local",
which does not make much sense inside of Cloudwatch.

Signed-off-by: Patrick Roy <[email protected]>
Before emitting raw data, we only submitted the average to display on
our dashboards. However, ever since submitting the raw data, we could
have just instructed cloudwatch to compute averages from that, without
needing to also submit the averages. So actually switch to doing that.

Also rename network_ab's "latency" metric to "ping_latency" for backward
compatibility.

Signed-off-by: Patrick Roy <[email protected]>
Because its a waste of comptue resources to run it if only markdown
files were changed.

Signed-off-by: Patrick Roy <[email protected]>
We want to call the one in host_tools.metrics so that the properties are
correctly set.

Signed-off-by: Patrick Roy <[email protected]>
The raw time series emitted from iperf3 include data points from the
warmup period. Since we do our A/B-Test on this entire time series, we
should exclude them.

Signed-off-by: Patrick Roy <[email protected]>
@roypat roypat added the Status: Awaiting review Indicates that a pull request is ready to be reviewed label Oct 2, 2023
@roypat roypat merged commit cd27794 into firecracker-microvm:main Oct 2, 2023
5 checks passed
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
Status: Awaiting review Indicates that a pull request is ready to be reviewed
Projects
None yet
Development

Successfully merging this pull request may close these issues.

3 participants