When virtio_fs_cache is set to none, the mount options for the folder
inside the guest should not contain the dax option else it leads to
invalid address errors and a crash of the daemon on the host.
Fixes: #1907
Signed-off-by: Ganesh Maharaj Mahalingam <ganesh.mahalingam@intel.com>
kata-agent is moved to /sbin/init when AGENT_INIT is yes. Check
/sbin/init exists and it's not systemd
fixes#322
Signed-off-by: Julio Montes <julio.montes@intel.com>
in create sandbox, if process error, should remove network without judge
NetNsCreated is true, since network is created by kata and should be
removed by kata, and network.Remove has judged if need to delete netns
depend on NetNsCreated
Fixes: #1920
Signed-off-by: Ace-Tang <aceapril@126.com>
Use .ci/setup.sh from tests repo, to make sure spell check packages are
correctly installed for the OS under test.
Signed-off-by: Marco Vedovati <mvedovati@suse.com>
While calculating directory size in imagebuild script, we shoul use
`df -BM` instead of `df -h` to keep Mb format.
Fixes#336
image-builder: get mount directory size with Mb format
Signed-off-by: ClarkLee <clarklee1992@hotmail.com>
Add the option to build image and initrd using dracut.
Fixes: #311
Suggested-by: Stefan Hajnoczi <stefanha@redhat.com>
Signed-off-by: Marco Vedovati <mvedovati@suse.com>
CPU topology has changed in QEMU 4.1: socket > die > core > thread.
die option must be specified in order to hotplug CPUs on x86_64
Depends-on: github.com/kata-containers/packaging#657
fixes#1913
Signed-off-by: Julio Montes <julio.montes@intel.com>
Shortlog:
68cdf64 test: add cpu topology tests
e0cf9d5 qmp: add checks for the CPU toplogy
a5c1190 qemu: support x86 SMP die
Signed-off-by: Julio Montes <julio.montes@intel.com>
Update top-level vendor with
"dep ensure add github.com/blang/semver@3.6.1"
dep check should not succeed.
Signed-off-by: Archana Shinde <archana.m.shinde@intel.com>
It is not really recommended to have nested vendor directories.
dep does not work well with nested directories:
https://github.com/golang/dep/issues/985
Recommendation is to use flatten the vendor directories.
Hence remove any nested vendor directories.
Fixes#1909
Signed-off-by: Archana Shinde <archana.m.shinde@intel.com>
For some reason, busybox image crashed on fedora 30 rootfs on aarch64.
For now, we will switch back to use fedora 28.
Fixes: #334
Signed-off-by: Penny Zheng <penny.zheng@arm.com>
Fixes#803
Merge "hypervisor.json" into "persist.json", so the new store can take
care of hypervisor data now.
Signed-off-by: Wei Zhang <weizhang555.zw@gmail.com>
For one thing, it is container specific resource so it should not
be cleaned up by the agent. For another thing, we can make container
stop to force cleanup these host mountpoints regardless of hypervisor
and agent liveness.
Signed-off-by: Peng Tao <bergwolf@hyper.sh>
Then we can check hypervisor liveness in those cases to avoid long
timeout when connecting to the agent when hypervisor is dead.
For kata-agent, we still use the kata-proxy pid for the same purpose.
Signed-off-by: Peng Tao <bergwolf@hyper.sh>
Once we have found the container, we should never fail SIGKILL.
It is possible to fail to send SIGKILL because hypervisor might
be gone already. If we fail SIGKILL, upper layer cannot really
proceed to clean things up.
Also there is no need to save sandbox here as we did not change
any state.
Signed-off-by: Peng Tao <bergwolf@hyper.sh>
When force is true, ignore any guest related errors. This can
be used to stop a sandbox when hypervisor process is dead.
Signed-off-by: Peng Tao <bergwolf@hyper.sh>
Whenever we fail to connect, do not make any more attempts.
More attempts are possible during cleanup phase but we should
not try to connect any more there.
Signed-off-by: Peng Tao <bergwolf@hyper.sh>
When a container is paused and something goes terribly
wrong, we still need to be able to clean thing up. A paused
container should be able to transit to stopped state as well
so that we can delete it properly.
Signed-off-by: Peng Tao <bergwolf@hyper.sh>
This allows travis to use the go install script instead of having a
hard coded golang version at travis.yml
Fixes#1903
Signed-off-by: Gabriela Cervantes <gabriela.cervantes.tellez@intel.com>
Move some of the functions in rootfs.sh (generate_dockerfile,
detect_go_version) in scripts/lib.sh, to make those functions reusable
outside of rootfs.sh.
Signed-off-by: Marco Vedovati <mvedovati@suse.com>
