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

tests: fix some failures reported by oci-runtime-validation #16

Merged
merged 17 commits into from
Mar 8, 2019

Conversation

giuseppe
Copy link
Member

No description provided.

Copy link

@mrunalp mrunalp left a comment

Choose a reason for hiding this comment

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

lgtm

giuseppe added 5 commits March 1, 2019 09:18
it seems oci-runtime-validation is creating config files where rootfs
is set to '.'.

Signed-off-by: Giuseppe Scrivano <[email protected]>
Signed-off-by: Giuseppe Scrivano <[email protected]>
Signed-off-by: Giuseppe Scrivano <[email protected]>
the OCI specifications say that "bind" or "rbind" are specified in the
options.

Signed-off-by: Giuseppe Scrivano <[email protected]>
@giuseppe giuseppe force-pushed the fix-oci-validation branch from 24989bf to cc263e1 Compare March 1, 2019 13:14
@giuseppe giuseppe force-pushed the fix-oci-validation branch from cc263e1 to 0df41fc Compare March 1, 2019 14:17
@giuseppe
Copy link
Member Author

giuseppe commented Mar 1, 2019

@mrunalp @alban with these changes I get down to 20 failures, which are the same ones I see with runc on Fedora 29.

@alban are the oci-runtime-validation tests supposed to work with runc and get no failures? If so, what environment should I use?

@alban
Copy link
Contributor

alban commented Mar 1, 2019

@giuseppe I haven't tried oci-runtime-validation against runc recently, but last time I tried, there was still some tests that don't pass.

/cc @dongsupark

@dongsupark
Copy link

@alban @giuseppe Exactly.
When I ran oci-runtime-validation against runc (probably 1.0-rc6), I got at least 10 failed tests. I have never been able to get the tests passed.
One of the reasons is that a PR to fix a compliance issue in runc hooks was never merged. AFAIK there could be other compliance issues, which block the final 1.0 release of runc.

@giuseppe
Copy link
Member Author

giuseppe commented Mar 2, 2019

When I ran oci-runtime-validation against runc (probably 1.0-rc6), I got at least 10 failed tests. I have never been able to get the tests passed.

on what system are you running the tests? I've tried several times both on Ubuntu and Fedora and I never manage to get less than 20 failures with runc

This is a quite weird requirement in the OCI specs, but let's make
sure we honor it.

Signed-off-by: Giuseppe Scrivano <[email protected]>
@dongsupark
Copy link

Fedora 29.
Anyway the number of failed tests can vary, so it can be >= 20 on some systems.
So I believe you. :-)

giuseppe added 6 commits March 2, 2019 22:55
Signed-off-by: Giuseppe Scrivano <[email protected]>
this is a weird requirement from the OCI specs, but it makes the
oci-runtime-tools tests pass.

Signed-off-by: Giuseppe Scrivano <[email protected]>
@giuseppe
Copy link
Member Author

giuseppe commented Mar 3, 2019

I've done some more changes and now crun gets to 0 failures 😃 🎉 🎉 🎉

I needed to fix some tests in runtime-tools. They were probably never really used as they fail with runc. I've opened a PR here: opencontainers/runtime-tools#687

Some of the issues I was seeing were on my end. I've had to change the CPU type for the VM I was using so that 1gb hugetlb can be used and I've needed to add a IDE disk so the major:minor 8:0 is a valid device.

The PR for runtime-tools doesn't affect which is still at 11 failures.

@giuseppe
Copy link
Member Author

giuseppe commented Mar 4, 2019

/cc @vbatts

@vbatts
Copy link
Contributor

vbatts commented Mar 4, 2019

Oh nice

@giuseppe giuseppe force-pushed the fix-oci-validation branch from 2b49e20 to d9e3a43 Compare March 6, 2019 21:37
@giuseppe giuseppe force-pushed the fix-oci-validation branch 8 times, most recently from 4421727 to 268b435 Compare March 7, 2019 08:15
@giuseppe giuseppe force-pushed the fix-oci-validation branch from 268b435 to 30def0a Compare March 7, 2019 09:32
Copy link
Contributor

@vbatts vbatts left a comment

Choose a reason for hiding this comment

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

LGTM

@giuseppe
Copy link
Member Author

giuseppe commented Mar 7, 2019

yey tests are happy! Now I've enabled the oci-validation as part of the PR merge tests. Except some tests that fail or are flaky on Travis (but I promise they work on a F29 VM!)

@giuseppe giuseppe merged commit 30def0a into master Mar 8, 2019
@giuseppe giuseppe deleted the fix-oci-validation branch March 8, 2019 15:44
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.

5 participants