Reestablish conn with mixnode if err occurs #501
Conversation
Another CI failure appears :) I'll take a look |
It's still hard to debug since we cannot see logs from nodes, but CI seems to be failed because CI was busy at that time (similar as #494 and #492). I think this PR didn't affect to the CI failure because this PR is very simple and doesn't change the normal behavior. After all, I'll open another PR which helps us to see logs from nodes when CI integration tests fail. |
So, looks like issue is in windows right? Why windows why??? |
It seems it's not only in windows. Codecov in ubuntu also fails with the similar (but maybe different) symptom. I'm trying to debug it again before merging this. |
I've found that CIs on Windows are 2~3x slower than Ubuntu for some reason, even for the master branch: https://github.com/logos-co/nomos-node/actions/runs/6743301131/job/18330953687. So, I think the CI failures of this PR aren't related to the changes of this PR. We'll be able to be more confident about it with PR #508 soon (If any unexpected error isn't found in the logs, we could say the failure is due to the slow machine). |
Sounds good. I will try locally as I have windows myself and post results here. |
|
Windows tests succeed locally. So probably something going around the gh runner. |
youngjoon-lee
force-pushed
the
gh-artifacts
branch
from
c4c212a
to
b18e85e
Compare
youngjoon-lee
force-pushed
the
reestablish-conn-fixed
branch
from
2f637c5
to
d4bcd70
Compare
youngjoon-lee
force-pushed
the
reestablish-conn-fixed
branch
from
46bf96e
to
2f0a837
Compare
I'm reopening the PR #445, now that CI (integration test) is stable.
99% is the same as the PR #445, but I added an one-line commit: 3414c58. Please double check this, @al8n.
Since #445 has been already reviewed, I didn't assign many reviewers. But, pls feel free to have a look.