[flaky test] failure in cri-tools test: Container Mount Propagation

[allencloud]

Ⅰ. Issue Description

I found that in the cri-tools, we always met the flaky test

• Failure [1.371 seconds]
[k8s.io] Container Mount Propagation
/home/travis/gopath/src/github.com/kubernetes-sigs/cri-tools/pkg/framework/framework.go:73
  runtime should support mount propagation
  /home/travis/gopath/src/github.com/kubernetes-sigs/cri-tools/pkg/validate/container_linux.go:45
    mount with 'rshared' should support propagation from host to container and vice versa [It]
    /home/travis/gopath/src/github.com/kubernetes-sigs/cri-tools/pkg/validate/container_linux.go:94
    len(output) should not be zero.
    Expected
        <int>: 0
    not to be zero-valued
    /home/travis/gopath/src/github.com/kubernetes-sigs/cri-tools/pkg/validate/container_linux.go:111

Could you help to take a look at this? @starnop @YaoZengzeng

Ⅱ. Describe what happened

cri-tools fails.

Ⅲ. Describe what you expected to happen

cri-tools passes.

Ⅳ. How to reproduce it (as minimally and precisely as possible)

Ⅴ. Anything else we need to know?

Ⅵ. Environment:

• pouch version (use pouch version): master code
• OS (e.g. from /etc/os-release): travis-ci
• Kernel (e.g. uname -a):
• Install tools:
• Others:

[starnop]

At present, almost all flaky tests are caused by IO problem , and that's one of the it. :)

[allencloud]

I see. Could you file an issue to explicitly show the detailed plan for how to tackle the IO problems in the code base? @starnop

[starnop]

I see. Could you file an issue to explicitly show the detailed plan for how to tackle the IO problems in the code base? @starnop

@fuweid is leading the way, and a roadmap is needed which I'm looking forward to.

[allencloud]

Can we close this now? @starnop @fuweid
I have not seen cri-tools failure for a long time. 😄

[fuweid]

Since the IO issue is gone, we can close this.