Add support to capture kernel profiles during performance testing #38402
Conversation
This file contains hidden or 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
|
/scale-egw |
|
/net-perf-gke |
5d7b2ee to
19130f2
Compare
|
/test |
pchaigno
approved these changes
Mar 24, 2025
asauber
approved these changes
Mar 26, 2025
Automatically create the report directory (when specified) if not already existing, to ensure that we can appropriately store the results. Signed-off-by: Marco Iorio <marco.iorio@isovalent.com>
That port configured there is never accessed, and it doesn't seem to match the one used by netperf. Let's just remove it to prevent confusion. Related: 24caafb ("cli/perf: fix command and port of server deployment") Signed-off-by: Marco Iorio <marco.iorio@isovalent.com>
In preparation for the subsequent introduction of new types of pods, let's slightly rework the categorization by introducing an explicit role label. Signed-off-by: Marco Iorio <marco.iorio@isovalent.com>
Extend the connectivity perf command with an extra option to automatically capture kernel profiles during the execution of each performance test, for later inspection. The resulting profiles, saved into the result directory, can be converted into an interactive SVG file via (see [1] for more info): stackcollapse-perf.pl $file | flamegraph.pl > out.svg Note that the kernel profiles capture shall be enabled only when running against disposable nodes dedicated for this scope. Indeed, the capture makes use of privileged pods to escape into the underlying host, install the necessary additional software and eventually run the appropriate `perf` commands. `perf` needs to be installed on the host, rather than in a container, as the version is specific to the underlying kernel. [1]: https://github.com/brendangregg/FlameGraph Signed-off-by: Marco Iorio <marco.iorio@isovalent.com>
Enable the new Cilium CLI option to automatically capture kernel profiles during performance tests in the net-perf-gke workflow. Signed-off-by: Marco Iorio <marco.iorio@isovalent.com>
Enable the new Cilium CLI option to automatically capture kernel profiles during performance tests in the scale-egw workflow. Signed-off-by: Marco Iorio <marco.iorio@isovalent.com>
19130f2 to
3f3e85e
Compare
|
/test |
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.
Extend the connectivity perf command with an extra option to automatically capture kernel profiles during the execution of each performance test, for later inspection. The resulting profiles, saved into the result directory, can be converted into an interactive SVG file via (see [1] for more info):
Note that the kernel profiles capture shall be enabled only when running against disposable nodes dedicated for this scope. Indeed, the capture makes use of privileged pods to escape into the underlying host, install the necessary additional software and eventually run the appropriate
perfcommands.perfneeds to be installed on the host, rather than in a container, as the version is specific to the underlying kernel.Additionally, enable this new feature in the net-perf-gke and scale-egw workflows.
Currently builds on top of #38245
[1]: https://github.com/brendangregg/FlameGraph