When powering off the system, we want the ocid service, to shutdown
all containers running on the system so they can cleanup properly
This patch will cleanup all pods on poweroff.
The ocid-shutdown.service drops a file /var/run/ocid.shutdown when the system
is shutting down. The ocid-shutdown.service should only be executed at system
shutdown.
On bootup sequence should be
start ocid.service
start ocid-shutdown.service (This is a NO-OP)
On system shutdown
stop ocid-shutdown.service (Creates /var/run/ocid.shutdown)
stop ocid.service (Notices /var/run/ocid.service and stops all pods before exiting.)
Signed-off-by: Daniel J Walsh <dwalsh@redhat.com>
Now that we have support for split std{out,err}, make sure that execsync
will correctly handle the split stdio properly. In addition, extend the
ctr logging test to make sure that the regular container logging is also
split correctly. We can't test !terminal containers because we only have
a single console for both std{out,err}.
Signed-off-by: Aleksa Sarai <asarai@suse.de>
Now that conmon splits std{out,err} for !terminal containers, ExecSync
can parse that output to return the correct std{out,err} split to the
kubelet. Invalid log lines are ignored but complained about.
Signed-off-by: Aleksa Sarai <asarai@suse.de>
While it's not currently possible to do this for terminal=true
containers, for !terminal containers we can create separate pipes for
stdout and stderr, and then log them separately. This is required for
k8s's conformance tests.
Signed-off-by: Aleksa Sarai <asarai@suse.de>
It's 2017, let's not stick with C89 (also for some reason the Travis
environment has a different -std= default value than my local machine).
Signed-off-by: Aleksa Sarai <asarai@suse.de>
The CRI requires us to prepend (timestamp, stream) to every line of the
output, and it's quite likely (especially in the !terminal case) that we
will read more than one line of output in the read loop.
So, we need to write out each line separately with the prepended
timestamps. Doing this the simple way (the final part of the buffer is
written partially if it doesn't end in a newline) makes the code much
simpler, with the downside that if we ever switch to multiple streams
for output we'll have to rewrite parts of this.
In addition, drop the debugging output of cri-o for each chunk read so
we stop spamming stderr. We can do this now because 8a928d06e7
("oci: make ExecSync with ExitCode != 0 act properly") actually fixed
how ExecSync was being handled (especially in regards to this patch).
Fixes: 1dc4c87c93 ("conmon: add timestamps to logs")
Signed-off-by: Aleksa Sarai <asarai@suse.de>
Previously we returned an internal error result when a program had a
non-zero exit code, which was incorrect. Fix this as well as change the
tests to actually check the "ExitCode" response from ExecSync (rather
than expecting ocic-ctr to return an internal error).
Signed-off-by: Aleksa Sarai <asarai@suse.de>
Since we no longer fall back to the noop plugin when
CNI configuration files are missing, and since the default
sandbox_config.json test file is running without host
networking, we must install the bridge and loopback
configuration files by default for tests to pass.
Signed-off-by: Samuel Ortiz <sameo@linux.intel.com>
We need to support cases where InitCNI() is called before
any CNI configuration files have been installed. This is
for example happening when deploying a k8s cluster with kubeadm.
kubeadm will start the DNS pod and it is left to the caller to
pick a network overlay and create the corresponding pods, that
will typically install a CNI configuration file first.
Here we address that issue by doing 2 things:
- Not returning an error when the default CNI config files
directory is empty.
- If it is empty, we start a monitoring thread (fsnotify based)
that will synchronize the network configuration when a CNI
file is installed there.
Signed-off-by: Samuel Ortiz <sameo@linux.intel.com>
Fix the following upstream k8s's e2e-node test:
```
should be able to pull from private registry with secret [Conformance]
```
Signed-off-by: Antonio Murdaca <runcom@redhat.com>
fsnotify is needed by the new ocicni monitoring implementation.
As ocicni switched to logrus, glog is no longer needed.
Signed-off-by: Samuel Ortiz <sameo@linux.intel.com>
To be consistent with the rest of the CRI-O logs, and to be able
to set the ocicni verbosity, we convert it from glog to logrus.
Signed-off-by: Samuel Ortiz <sameo@linux.intel.com>
When RunPodSandbox fails after calling s.addSandbox(sb),
we're left with a sandbox in s.state.sandboxes while the
sandbox is not created.
We fix that by adding removeSandbox() to the deferred cleanup
call
Signed-off-by: Samuel Ortiz <sameo@linux.intel.com>
This patch prevents k8s's e2e_node tests from killing CRI-O (because of
a panic in marshaling nil responses). This will ensure tests keep
running and just logging the failure.
Signed-off-by: Antonio Murdaca <runcom@redhat.com>
As of containernetworking/cni commit 1b9caefba5670c59e0ccbf0b008d88da52a7d498,
the script to build has changed from 'build' to 'build.sh' which broke our
integration tests. Change the integration test Dockerfile to update this path,
and update the tutorial for good measure. Pin to current master to ensure this
breakage doesn't happen again.
Signed-off-by: Matthew Heon <mheon@redhat.com>
Because kubelet will create broken symlinks for logPath it is necessary
to remove those symlinks before we attempt to write to them. This is a
temporary workaround while the issue is fixed upstream.
Ref: https://issues.k8s.io/44043
Signed-off-by: Aleksa Sarai <asarai@suse.de>
The main purpose of these tests is to make sure that the log actually
contains output from the container. We don't test the timestamps or the
stream that's stated at the moment.
Signed-off-by: Aleksa Sarai <asarai@suse.de>
This is necessary, as otherwise ocid will use its own current directory
as a log_directory, which is not the best idea in the world. The same
applies for log_path.
Signed-off-by: Aleksa Sarai <asarai@suse.de>
CRI requires us to timestamp our logs line-by-line by specifying whether
the line came from std{out,err} and the time at which the log was
recieved. This is a preliminary implementation of said behaviour
(without explicit newline handling at the moment).
Signed-off-by: Mrunal Patel <mpatel@redhat.com>
Signed-off-by: Aleksa Sarai <asarai@suse.de>
While pipes have their downsides, it turns out that socketpair(2) will
break any program that tries to open /dev/std{out,err} for writing
(because they're symlinked to /proc/1/fd/{1,2} which will cause lots of
fun issues with sockets).
Signed-off-by: Mrunal Patel <mpatel@redhat.com>
Signed-off-by: Aleksa Sarai <asarai@suse.de>
This adds a very simple implementation of logging within conmon, where
every buffer read from the masterfd of the container is also written to
the log file (with errors during writing to the log file ignored).
Signed-off-by: Aleksa Sarai <asarai@suse.de>
The runtimeversion test was incorrectly written and would fail for no
good reason if setup_ocid happened to run a command that failed (even if
it was handled).
Signed-off-by: Aleksa Sarai <asarai@suse.de>
Interleaving asynchronous updates with pod or container creations can
lead to unrecoverable races and corruptions of the pod or container hash
tables. This is fixed by serializing update against pod or container
creation operations, while pod and container creation operations can
run in parallel.
Signed-off-by: Samuel Ortiz <sameo@linux.intel.com>
Now that the image package has fixes to support docker images v2s1,
we can remove our buildOCIProcessARgs() hack for empty image configs
and simplify this routine.
Signed-off-by: Samuel Ortiz <sameo@linux.intel.com>
Signed-off-by: Antonio Murdaca <runcom@redhat.com>
We want new sandboxes to be added to the sandbox hash table before
adding their ID to the pod Index registrar, in order to avoid potential
Update() races.
Signed-off-by: Samuel Ortiz <sameo@linux.intel.com>
