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

Clean up performance test #149

Merged
merged 1 commit into from
Sep 1, 2023
Merged

Clean up performance test #149

merged 1 commit into from
Sep 1, 2023

Conversation

crawford
Copy link
Contributor

The spirit of the test remains the same - test the throughput of a simple webserver, which has to do some amount of computation, within an enclave - but the implementation has been simplified. The documentation and scripts have also been cleaned up to simplify the setup and execution of the test.

robszumski
robszumski previously approved these changes Aug 30, 2023
Copy link
Contributor

@robszumski robszumski left a comment

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

LGTM. I assume the time for encryption was negligible, hence the removal?

The spirit of the test remains the same - test the throughput of a
simple webserver, which has to do some amount of computation, within
an enclave - but the implementation has been simplified. The
documentation and scripts have also been cleaned up to simplify the
setup and execution of the test.

During testing, I was able to achieve almost 500 requests per second
for both two and four vCPUs. This is mainly due to the fact that the
ramp-up time is set to 20 seconds (the full test runs take 24
seconds). As I increase the amount of busy-looping per request, I
start to see more errors (and it's not immediately obvious what
constitutes an error). I'm leaving the test and the result as is and
will follow up with a better benchmark implementation.
@crawford
Copy link
Contributor Author

crawford commented Sep 1, 2023

I replaced the encryption with a busy loop to simplify the implementation (and to better control how long each request takes). We aren't benchmarking Golang's encryption implementation, so I wanted to remove that variable from the picture.

Copy link
Contributor

@robszumski robszumski left a comment

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

LGTM, nice busy loop in there

@crawford crawford merged commit 2173eaa into edgebitio:main Sep 1, 2023
1 check passed
@crawford crawford deleted the perf branch September 1, 2023 20:16
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
None yet
Projects
None yet
Development

Successfully merging this pull request may close these issues.

2 participants