The create/start split left checkpoint/restore broken in that the create
side was calling runc restore, which fully restores the process to
operation, leaving a call to `runc start` as an error (process already
started). This patch skips process.Start as it is an unnecessary (and
wrong) step for a restored checkpoint.
Signed-off-by: Phil Estes <estesp@linux.vnet.ibm.com>
That comment didn't seem to match the reality and confused me a bit. The
shim takes 3 arguments and the cwd is the containerd state directory.
I'm guessing it's a left over from the containerd split.
Signed-off-by: Damien Lespiau <damien.lespiau@intel.com>
* gitignore: Use absolute paths
git status will only match the ignored path from the root of the
repository, a very slightly more accurate way to ignore those files.
Signed-off-by: Damien Lespiau <damien.lespiau@intel.com>
* gitignore: Ignore the /output directory
Test artifacts are now generated in /output, so update the .gitignore
file accordingly.
Signed-off-by: Damien Lespiau <damien.lespiau@intel.com>
Sigchld reaper has been removed from containerd procss in
847690583f, so the comment
is not need any more.
Signed-off-by: Lei Jitang <leijitang@huawei.com>
To improve the output of module logging, we now include a path that
represents the modules that a context has passed through. This makes it
easy to tell if a log line came from a particular module hierarchy.
A helper function, `WithModule`, provides an easy way to mark a context
with a module and inject that module into the logger.
Signed-off-by: Stephen J Day <stephen.day@docker.com>
* Sync process.State() with the matching events
Signed-off-by: Kenfe-Mickael Laventure <mickael.laventure@gmail.com>
* Allow requesting events for a specific container
Signed-off-by: Kenfe-Mickael Laventure <mickael.laventure@gmail.com>
* Sync container state retrieval with other events
Signed-off-by: Kenfe-Mickael Laventure <mickael.laventure@gmail.com>
* Let containerd take care of calling runtime delete on exit
Signed-off-by: Kenfe-Mickael Laventure <mickael.laventure@gmail.com>
* Take care of possible race in TestBusyboxTopExecTopKillInit
Signed-off-by: Kenfe-Mickael Laventure <mickael.laventure@gmail.com>
This fixes a sync issue when the containerd api returns after a
container has started. It fixes it by calling the runtime start inside
containerd after the oom handler has been setup.
Signed-off-by: Michael Crosby <crosbymichael@gmail.com>
Make sure that if we get an error from start it is returned and that we
reap the container's process from create before existing if we were to
send a kill signal.
Signed-off-by: Michael Crosby <crosbymichael@gmail.com>
* Micro benchmarks: use container.Runtime to kill container
Signed-off-by: Julio Montes <julio.montes@intel.com>
* Micro benchmarks: add support for multiples runtimes
Signed-off-by: Julio Montes <julio.montes@intel.com>
This will prevent from leaving hung processes around if there's a bug
in the runtime.
Signed-off-by: Kenfe-Mickael Laventure <mickael.laventure@gmail.com>
This is needed to get the new `--empty-ns` flag for the restore
operation.
NOTE: I've also updated the vendor.sh script even though this doesn't
pull any new code in
Signed-off-by: Kenfe-Mickael Laventure <mickael.laventure@gmail.com>
