Add nightly CI check against latest substrate. #335
Conversation
| uses: actions-rs/cargo@v1.0.3 | ||
| with: | ||
| command: test | ||
| args: --all-targets --workspace |
There was a problem hiding this comment.
maybe do something like: https://github.heygears.community/t/create-an-issue-case-the-workflow-fails/17247
There was a problem hiding this comment.
Ooh, great idea! I'll look at that :)
|
I realised that the reason that the doc step was not finishing was because the Edit: Making sure to kill the substrate process on panic (with a custom drop impl) seems to have resolved the issue of panics blocking completion) |
…ic, to prevent blocking
jsdw
force-pushed
the
jsdw-nightly-ci
branch
from
847f624 to
a9430a9
Compare
jsdw
changed the title
|
Just FYI: this issue was opened to address the question of "which substrate binary to pin to for PR runs": https://github.com/paritytech/ci_cd/issues/272 |
* Add nightly check and use specific version (currently the same) in PR checks * wrong var name, whoops * Create GH issue if nightly run fails * need tokio feature for RpcClient; match dependency in root Cargo.toml. * Ensure substrate command is properly cleaned up in the event of a panic, to prevent blocking * increase retry limit again * fmt * Tweak text for nightly-run-failed issue template * Run nightly-CI every 5mins for testing * I clearly dont use cron enough * Put back cron timings to 8am daily * break a test * revert nightly workflow again (doesnt run off master) and testing rust workflow * more testing.. * undo testing things; seems to work * tidy up; substrate url at top of both CI files
Rel #334