-
Notifications
You must be signed in to change notification settings - Fork 14
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
Add support for Azure Cleanroom based CCF #215
Conversation
734367f
to
7c8b6a6
Compare
. TMP: Use permitted sub . . . Avoid races setting up ccf providers . . . Fix permissions on setup.sh Pre-start jwt issuer to prevent races Run tests sequentially Fix JWT test
3f302a7
to
04159d1
Compare
This reverts commit 4c78bea.
This reverts commit d15d9f9.
Run `az-cleanroom-aci` tests in parallel
on: | ||
workflow_dispatch: | ||
schedule: | ||
- cron: "0 0 * * *" |
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
Doing this every night is problematic. Could we do the cleanup on Sunday?
Scenario:
We have large builds/tests. Start a full build/test before shutting down the day to inspect the next morning.
Or maybe we can do cleanup for containers older than e.g. 2 days
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
Nightly cleanup is how all of our other repos work, including privacy-sandbox-dev which has about an order of magnitude more expensive builds/tests than this repo. What are you envisioning taking overnight and relying on having something deployed in ACI for that whole time?
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
This is a plead for not doing nightly cleanups for any of our repos. At least wait one day so if you come back in the morning you can still inspect the failing tests
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
The cleanup is before the nightly testing, so you can inspect the resources deployed by a given failing run for a full work day before it's cleaned up. And of course the logs in actions itself should have a reasonable amount of information to be useful too and that lasts much longer
source ./scripts/jwt_issuer/up.sh | ||
pytest -sv \ | ||
test/system-test/test_${{ inputs.endpoint }}.py \ | ||
-n ${{ matrix.config.threads }} |
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
cool. Can we use -n on our other CI's?
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
If by other CI's you mean using the sandbox_local
environment, no because it's currently got hardcoded endpoint ports so if you tried to run multiple concurrently they would clash.
@@ -1,5 +1,5 @@ | |||
# Core Environment Variables | |||
export SUBSCRIPTION=85c61f94-8912-4e82-900e-6ab44de9bdf8 | |||
export SUBSCRIPTION=7ca35580-fc67-469c-91a7-68b38569ca6e |
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
Shouldn't this be a secret?
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
Why? It's just an ID you can't do anything with it without sufficient role based access
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
OK fair
Motivation
This allows us to test against a production-like mCCF replacement, and will allow us to validate it against a realistic use case for us
Changes
scripts/az-cleanroom-aci
dir which installs the extension and brings up the networkconftest.py
for deploying the networkup.sh
(maybeaz cleanroom ccf network down
?)Current compromises:
az-cleanroom-aci
tests in parallel #224 will address this but it's currently blocked onaz cleanroom
so keeping separate so CI on this PR passesaz cleanroom
doesn't specify a subject alt name so the service cert isn't usable, so I suggest for now we remove SSL cert verification in the endpoint scripts (i.e. add-k
). I will raise an issue with them and when it's fixed we can remove-k