cmd actions bench & fmt #406
Conversation
There was a problem hiding this comment.
Good work, thanks! I mostly skimmed it but i think this is one of the things where we just need to experiment a bit.
Only point is that there should be a commands.md or bot.md with some overview of how the bot works, what commands are available and how the components make it work.
The weight-generation file seems to be specific already.
Co-authored-by: Oliver Tale-Yazdi <[email protected]>
|
Review required! Latest push from author must always be reviewed |
1 similar comment
|
Review required! Latest push from author must always be reviewed |
.github/workflows/cmd.yml
Outdated
|
|
||
| jobs: | ||
| acknowledge: | ||
| if: ${{ startsWith(github.event.comment.body, '/cmd') }} |
There was a problem hiding this comment.
yep good point, will add a similar check as for /merge
and cmd.md guide
thanks
|
I don't know what's the plan for the self hosted runners but for ORML we are using github webhook to create new AWS EC2 instance upon request and autoscale it down when idle. In that way we don't need to run idle instance and able to scale up as many runners as needed. |
| CMD: ${{ steps.get-pr-comment.outputs.group2 }} # to avoid "" around the command | ||
| run: | | ||
| echo 'help<<EOF' >> $GITHUB_OUTPUT | ||
| python3 .github/scripts/cmd/cmd.py $CMD >> $GITHUB_OUTPUT |
There was a problem hiding this comment.
This line doesn't look safe, I'll write in Element.
There was a problem hiding this comment.
Improved regex which prevents passing arbitrary commands + check for fellows
mordamax
force-pushed
the
mak-testing-benchmarks-2
branch
2 times, most recently
from
fbdf2d6
to
0bed466
Compare
mordamax
force-pushed
the
mak-testing-benchmarks-2
branch
from
2d4f52b
to
b2d273d
Compare
|
While this is being prepared, what is the standard way to generate weight files? I suppose I need access to a 'standard machine'? Asking for #364 |
if it waits until early next week, i'd recommend to wait a bit. @tugytur was going to spawn a runner as soon as @bkchr provides details, hope today or early next week 🤞 If not, then you can follow this guidance https://github.com/polkadot-fellows/runtimes/blob/main/docs/weight-generation.md and find |
|
Not super rushed from my side, but thanks for the context :) |
|
/merge |
|
Enabled Available commands
For more information see the documentation |
fellowship-merge-bot
bot
merged commit 16d2db7
into
polkadot-fellows:main
Closes #128
Closes https://github.com/paritytech/infrastructure/issues/41
Adds ability to run some standard helper commands, for now only for benchmarks and fmt (with taplo)
It will commit the result of operation back into PR with some reports and/or links to logs
bench uses
gitrun-001runner which corresponds to Validator Reference HardwareExamples (screenshots):
Benchmarks
Bench results with Subweight report
/cmd [command] --clean - removing bot's and author's command runs to clean up PR
/cmd [command] --help - outputs the usage instructions