Today in agent watchers, when we copy files/symlinks
or create directories, the ownership of the source path
is not preserved which can lead to permission issues.
In copy, ensure that we do a chown of the source path
uid/gid to the destination file/symlink after copy to
ensure that ownership matches the source ownership.
fs::copy() takes care of setting the permissions.
For directory creation, ensure that we set the
permissions of the created directory to the source
directory permissions and also perform a chown of the
source path uid/gid to ensure directory ownership
and permissions matches to the source.
Fixes: #4188
Signed-off-by: Yibo Zhuang <yibzhuang@gmail.com>
(cherry picked from commit 70eda2fa6c)
Signed-off-by: Fabiano Fidêncio <fabiano.fidencio@intel.com>
We're currently hitting a race condition on the Cloud Hypervisor's
driver code when quickly removing and adding a block device.
This happens because the device removal is an asynchronous operation,
and we currently do *not* monitor events coming from Cloud Hypervisor to
know when the device was actually removed. Together with this, the
sandbox code doesn't know about that and when a new device is attached
it'll quickly assign what may be the very same ID to the new device,
leading to the Cloud Hypervisor's driver trying to hotplug a device with
the very same ID of the device that was not yet removed.
This is, in a nutshell, why the tests with Cloud Hypervisor and
devmapper have been failing every now and then.
The workaround taken to solve the issue is basically *not* passing down
the device ID to Cloud Hypervisor and simply letting Cloud Hypervisor
itself generate those, as Cloud Hypervisor does it in a manner that
avoids such conflicts. With this addition we have then to keep a map of
the device ID and the Cloud Hypervisor's generated ID, so we can
properly remove the device.
This workaround will probably stay for a while, at least till someone
has enough cycles to implement a way to watch the device removal event
and then properly act on that. Spoiler alert, this will be a complex
change that may not even be worth it considering the race can be avoided
with this commit.
Fixes: #4196
Signed-off-by: Fabiano Fidêncio <fabiano.fidencio@intel.com>
(cherry picked from commit 33a8b70558)
Signed-off-by: Fabiano Fidêncio <fabiano.fidencio@intel.com>
This allows to get released binaries again.
Fixes: #4151
Signed-off-by: Greg Kurz <groug@kaod.org>
(cherry picked from commit b658dccc5f)
Signed-off-by: Fabiano Fidêncio <fabiano.fidencio@intel.com>
If we fail to download the clh binary, we fall-back to build from source.
Unfortunately, `pull_clh_released_binary()` leaves a `cloud_hypervisor`
directory behind, which causes `build_clh_from_source()` not to clone
the git repo:
[ -d "${repo_dir}" ] || git clone "${cloud_hypervisor_repo}"
When building from a kata-containers git repo, the subsequent calls
to `git` in this function thus apply to the kata-containers repo and
eventually fail, e.g.:
+ git checkout v23.0
error: pathspec 'v23.0' did not match any file(s) known to git
It doesn't quite make sense actually to keep an existing directory the
content of which is arbitrary when we want to it to contain a specific
version of clh. Just remove it instead.
Fixes: #4151
Signed-off-by: Greg Kurz <groug@kaod.org>
(cherry picked from commit afbd60da27)
Signed-off-by: Fabiano Fidêncio <fabiano.fidencio@intel.com>
The type of MemorySwappiness in runtime is uint64, and the type of swappiness in agent is int64,
if we set max uint64 in runtime and pass it to agent, the value will be equal to -1. We should
modify the type of swappiness to u64
Fixes: #4123
Signed-off-by: holyfei <yangfeiyu20092010@163.com>
(cherry picked from commit 0239502781)
Signed-off-by: Fabiano Fidêncio <fabiano.fidencio@intel.com>
Currently EnableMockTesting() takes no arguments and will always place the
mock storage in the fixed location /tmp/vc/mockfs. This means that one
test run can interfere with the next one if anything isn't cleaned up
(and there are other bugs which means that happens). If if those were
fixed this would allow developers testing on the same machine to interfere
with each other.
So, allow the mockfs to be placed at an arbitrary place given as a
parameter to EnableMockTesting(). In TestMain() we place it under our
existing temporary directory, so we don't need any additional cleanup just
for the mockfs.
fixes#4140
Signed-off-by: David Gibson <david@gibson.dropbear.id.au>
(cherry picked from commit 1b931f4203)
Signed-off-by: Fabiano Fidêncio <fabiano.fidencio@intel.com>
Currently MockFSInit always creates the mockfs at the fixed path
/tmp/vc/mockfs. This change allows it to be initialized at any path
given as a parameter. This allows the tests in fs_test.go to be
simplified, because the by using a temporary directory from
t.TempDir(), which is automatically cleaned up, we don't need to
manually trigger initTestDir() (which is misnamed, it's actually a
cleanup function).
For now we still use the fixed path when auto-creating the mockfs in
MockAutoInit(), but we'll change that later.
Signed-off-by: David Gibson <david@gibson.dropbear.id.au>
(cherry picked from commit ef6d54a781)
Signed-off-by: Fabiano Fidêncio <fabiano.fidencio@intel.com>
virtcontainers/persist/fs/mockfs.go defines a mock filesystem type for
testing. A global variable in virtcontainers/persist/manager.go is used to
force use of the mock fs rather than a normal one.
This patch moves the global, and the EnableMockTesting() function which
sets it into mockfs.go. This is slightly cleaner to begin with, and will
allow some further enhancements.
Signed-off-by: David Gibson <david@gibson.dropbear.id.au>
(cherry picked from commit 5d8438e939)
Signed-off-by: Fabiano Fidêncio <fabiano.fidencio@intel.com>
storagePathSuffix defines the file path suffix - "vc" - used for
Kata's persistent storage information, as a private constant. We
duplicate this information in fc.go which also needs it.
Export it from fs.go instead, so it can be used in fc.go.
Signed-off-by: David Gibson <david@gibson.dropbear.id.au>
(cherry picked from commit 963d03ea8a)
Signed-off-by: Fabiano Fidêncio <fabiano.fidencio@intel.com>
A number of unit tests under virtcontainers/factory use
MockStorageRootPath() as a general purpose temporary directory. This
doesn't make sense: the mockfs driver isn't even in use here since we only
call EnableMockTesting for the pase virtcontainers package, not the
subpackages.
Instead use t.TempDir() which is for exactly this purpose. As a bonus it
also handles the cleanup, so we don't need MockStorageDestroy any more.
Signed-off-by: David Gibson <david@gibson.dropbear.id.au>
(cherry picked from commit 1719a8b491)
Signed-off-by: Fabiano Fidêncio <fabiano.fidencio@intel.com>
There are several tests in mount_test.go which perform a sample bind
mount. These need a corresponding unmount to clean up afterwards or
attempting to delete the temporary files will fail due to the existing
mountpoint. Most of them had such an unmount, but
TestBindMountInvalidPgtypes was missing one.
In addition, the existing unmounts where done inconsistently - one was
simply inline (so wouldn't be executed if the test fails too early) and one
is a defer. Change them all to use the t.Cleanup mechanism.
For the dummy mountpoint files, rather than cleaning them up after the
test, the tests were removing them at the beginning of the test. That
stops the test being messed up by a previous run, but messily. Since
these are created in a private temporary directory anyway, if there's
something already there, that indicates a problem we shouldn't ignore.
In fact we don't need to explicitly remove these at all - they'll be
removed along with the rest of the private temporary directory.
Signed-off-by: David Gibson <david@gibson.dropbear.id.au>
(cherry picked from commit bec59f9e39)
Signed-off-by: Fabiano Fidêncio <fabiano.fidencio@intel.com>
The tests in hook_test.go run a mock hook binary, which does some debug
logging to /tmp/mock_hook.log. Currently we don't clean up those logs
when the tests are done. Use a test cleanup function to do this.
Signed-off-by: David Gibson <david@gibson.dropbear.id.au>
(cherry picked from commit f7ba21c86f)
Signed-off-by: Fabiano Fidêncio <fabiano.fidencio@intel.com>
SetupOCIConfigFile creates a temporary directory with os.MkDirTemp(). This
means the callers need to register a deferred function to remove it again.
At least one of them was commented out meaning that a /temp/katatest-
directory was leftover after the unit tests ran.
Change to using t.TempDir() which as well as better matching other parts of
the tests means the testing framework will handle cleaning it up.
Signed-off-by: David Gibson <david@gibson.dropbear.id.au>
(cherry picked from commit 90b2f5b776)
Signed-off-by: Fabiano Fidêncio <fabiano.fidencio@intel.com>
Several tests in kata_agent_test.go create /tmp/mountPoint as a dummy
directory to mount. This is not cleaned up after the test. Although it
is in /tmp, that's still a little messy and can be confusing to a user.
In addition, because it uses the same name every time, it allows for one
run of the test to interfere with the next.
Use the built in t.TempDir() to use an automatically named and deleted
temporary directory instead.
Signed-off-by: David Gibson <david@gibson.dropbear.id.au>
(cherry picked from commit 2eeb5dc223)
Signed-off-by: Fabiano Fidêncio <fabiano.fidencio@intel.com>
Add some links to rendered webpages for better user experience,
let users can jump to pages only by clicking links in browsers.
Fixes: #4061
Signed-off-by: bin <bin@hyper.sh>
(cherry picked from commit f8cc5d1ad8)
Signed-off-by: Fabiano Fidêncio <fabiano.fidencio@intel.com>
Adding two functions set_ownership and
recursive_ownership_change to support changing group id
ownership for a mounted volume.
The set_ownership will be called in common_storage_handler
after mount_storage performs the mount for the volume.
set_ownership will be a noop if the FSGroup field in the
Storage struct is not set which indicates no chown will be
performed. If FSGroup field is specified, then it will
perform the recursive walk of the mounted volume path to
change ownership of all files and directories to the
desired group id. It will also configure the SetGid bit
so that files created the directory will have group
following parent directory group.
If the fsGroupChangePolicy is on root mismatch,
then the group ownership will be skipped if the root
directory group id alreasy matches the desired group
id and if the SetGid bit is also set on the root directory.
This is the same behavior as what
Kubelet does today when performing the recursive walk
to change ownership.
Fixes#4018
Signed-off-by: Yibo Zhuang <yibzhuang@gmail.com>
(cherry picked from commit 92c00c7e84)
Signed-off-by: Fabiano Fidêncio <fabiano.fidencio@intel.com>
This change adds two fields to the Storage pb
FSGroup which is a group id that the runtime
specifies to indicate to the agent to perform a
chown of the mounted volume to the specified
group id after mounting is complete in the guest.
FSGroupChangePolicy which is a policy to indicate
whether to always perform the group id ownership
change or only if the root directory group id
does not match with the desired group id.
These two fields will allow CSI plugins to indicate
to Kata that after the block device is mounted in
the guest, group id ownership change should be performed
on that volume.
Fixes#4018
Signed-off-by: Yibo Zhuang <yibzhuang@gmail.com>
(cherry picked from commit 6a47b82c81)
Signed-off-by: Fabiano Fidêncio <fabiano.fidencio@intel.com>
The fsGroup will be specified by the fsGroup key in
the direct-assign mountinfo metadate field.
This will be set when invoking the kata-runtime
binary and providing the key, value pair in the metadata
field. Similarly, the fsGroupChangePolicy will also
be provided in the mountinfo metadate field.
Adding an extra fields FsGroup and FSGroupChangePolicy
in the Mount construct for container mount which will
be populated when creating block devices by parsing
out the mountInfo.json.
And in handleDeviceBlockVolume of the kata-agent client,
it checks if the mount FSGroup is not nil, which
indicates that fsGroup change is required in the guest,
and will provide the FSGroup field in the protobuf to
pass the value to the agent.
Fixes#4018
Signed-off-by: Yibo Zhuang <yibzhuang@gmail.com>
(cherry picked from commit 532d53977e)
Signed-off-by: Fabiano Fidêncio <fabiano.fidencio@intel.com>
The scanner reads nothing from viriofsd stderr pipe, because param
'--syslog' rediercts stderr to syslog. So there is no need to write
scanner.Text() to kata log
Fixes: #4063
Signed-off-by: Zhuoyu Tie <tiezhuoyu@outlook.com>
(cherry picked from commit 6e79042aa0)
Signed-off-by: Fabiano Fidêncio <fabiano.fidencio@intel.com>
kata-monitor allows to get data profiles from the kata shim
instances running on the same node by acting as a proxy
(e.g., http://$NODE_ADDRESS:8090/debug/pprof/?sandbox=$MYSANDBOXID).
In order to proxy the requests and the responses to the right shim,
kata-monitor requires to pass the sandbox id via a query string in the
url.
The profiling index page proxied by kata-monitor contains the link to all
the data profiles available. All the links anyway do not contain the
sandbox id included in the request: the links result then broken when
accessed through kata-monitor.
This happens because the profiling index page comes from the kata shim,
which will not include the query string provided in the http request.
Let's add on-the-fly the sandbox id in each href tag returned by the kata
shim index page before providing the proxied page.
Fixes: #4054
Signed-off-by: Francesco Giudici <fgiudici@redhat.com>
(cherry picked from commit 86977ff780)
Signed-off-by: Fabiano Fidêncio <fabiano.fidencio@intel.com>
During container exit, the agent tries to remove all the mount point directories,
which can fail if it's a readonly filesytem (e.g. device mapper). This commit ignores
the removal failure and logs a warning message.
Fixes: #4043
Signed-off-by: Feng Wang <feng.wang@databricks.com>
(cherry picked from commit aabcebbf58)
Signed-off-by: Fabiano Fidêncio <fabiano.fidencio@intel.com>
`make kata-tarball` relies on `kata-deploy-binaries.sh -s` which
silently ignores errors, and you may end up with an incomplete
tarball without noticing it because `make`'s exit status is 0.
`kata-deploy-binaries.sh` does set the `errexit` option and all the
code in the script seems to assume that since it doesn't do error
checking. Unfortunately, bash automatically disables `errexit` when
calling a function from a conditional pipeline, like done in the `-s`
case:
if [ "${silent}" == true ]; then
if ! handle_build "${t}" &>"$log_file"; then
^^^^^^
this disables `errexit`
and `handle_build` ends with a `tar tvf` that always succeeds.
Adding error checking all over the place isn't really an option
as it would seriously obfuscate the code. Drop the conditional
pipeline instead and print the final error message from a `trap`
handler on the special ERR signal. This requires the `errtrace`
option as `trap`s aren't propagated to functions by default.
Since all outputs of `handle_build` are redirected to the build
log file, some file descriptor duplication magic is needed for
the handler to be able to write to the orignal stdout and stderr.
Fixes#3757
Signed-off-by: Greg Kurz <groug@kaod.org>
(cherry picked from commit a779e19bee)
Signed-off-by: Fabiano Fidêncio <fabiano.fidencio@intel.com>
'make kata-tarball' sometimes fails early with:
cp: cannot create regular file '[...]/tools/packaging/kata-deploy/local-build/dockerbuild/install_yq.sh': File exists
This happens because all assets are built in parallel using the same
`kata-deploy-binaries-in-docker.sh` script, and thus all try to copy
the `install_yq.sh` script to the same location with the `cp` command.
This is a well known race condition that cannot be avoided without
serialization of `cp` invocations.
Move the copying of `install_yq.sh` to a separate script and ensure
it is called *before* parallel builds. Make the presence of the copy
a prerequisite for each sub-build so that they still can be triggered
individually. Update the GH release workflow to also call this script
before calling `kata-deploy-binaries-in-docker.sh`.
Fixes#3756
Signed-off-by: Greg Kurz <groug@kaod.org>
(cherry picked from commit 154c8b03d3)
Signed-off-by: Fabiano Fidêncio <fabiano.fidencio@intel.com>
NO_TTY configured whether to add the -t option to docker run. It makes no
sense for the caller to configure this, since whether you need it depends
on the commands you're running. Since the point here is to run
non-interactive build scripts, we don't need -t, or -i either.
Signed-off-by: David Gibson <david@gibson.dropbear.id.au>
Signed-off-by: Greg Kurz <groug@kaod.org>
(cherry picked from commit 1ed7da8fc7)
Signed-off-by: Fabiano Fidêncio <fabiano.fidencio@intel.com>
This directory consists entirely of files built during a make kata-tarball,
so it should not be committed to the tree. A symbolic link to this directory
might be created during 'make tarball', ignore it as well.
Signed-off-by: David Gibson <david@gibson.dropbear.id.au>
[greg: - rearranged the subject to make the subsystem checker happy
- also ignore the symbolic link created by
`kata-deploy-binaries-in-docker.sh`]
Signed-off-by: Greg Kurz <groug@kaod.org>
(cherry picked from commit bad859d2f8)
Signed-off-by: Fabiano Fidêncio <fabiano.fidencio@intel.com>
This release changes Docker images repository from DockerHub to Amazon
ECR. This resolves the `You have reached your pull rate limit` error
when building the firecracker tarball.
Fixes#4001
Signed-off-by: Greg Kurz <groug@kaod.org>
(cherry picked from commit 0d5f80b803)
Signed-off-by: Fabiano Fidêncio <fabiano.fidencio@intel.com>
The hugepages resources claimed by containers should be limited
by cgroup in the guest OS.
Fixes: #3695
Signed-off-by: Miao Xia <xia.miao1@zte.com.cn>
(cherry picked from commit a2f5c1768e)
Signed-off-by: Fabiano Fidêncio <fabiano.fidencio@intel.com>
allocate_hugepages() writes to the kernel sysfs file to allocate hugepages
in the Kata VM. However, even if the write succeeds, it's not certain that
the kernel will actually be able to allocate as many hugepages as we
requested.
This patch reads back the file after writing it to check if we were able to
allocate all the required hugepages.
fixes#3816
Signed-off-by: David Gibson <david@gibson.dropbear.id.au>
(cherry picked from commit 42e35505b0)
Signed-off-by: Fabiano Fidêncio <fabiano.fidencio@intel.com>
allocate_hugepages() constructs the path for the sysfs directory containing
hugepage configuration, then attempts to create this directory if it does
not exist.
This doesn't make sense: sysfs is a view into kernel configuration, if the
kernel has support for the hugepage size, the directory will already be
there, if it doesn't, trying to create it won't help.
For the same reason, attempting to create the "nr_hugepages" file
itself is pointless, so there's no reason to call
OpenOptions::create(true).
Signed-off-by: David Gibson <david@gibson.dropbear.id.au>
(cherry picked from commit 608e003abc)
Signed-off-by: Fabiano Fidêncio <fabiano.fidencio@intel.com>
Highlights from the Cloud Hypervisor release v23.0: 1) vDPA Support; 2)
Updated OS Support list (Jammy 22.04 added with EOLed versions removed);
3) AArch64 Memory Map Improvements; 4) AMX Support; 5) Bug Fixes;
Details can be found: https://github.com/cloud-hypervisor/cloud-hypervisor/releases/tag/v23.0Fixes: #4101
Signed-off-by: Bo Chen <chen.bo@intel.com>
(cherry picked from commit 6012c19707)
This is to avoid accidentally deleting multiple volumes.
Fixes#4020
Signed-off-by: Feng Wang <feng.wang@databricks.com>
(cherry picked from commit 354cd3b9b6)
- stable-2.4 | agent: fix container stop error with signal SIGRTMIN+3
- stable-2.4 | kata-monitor: fix duplicated output when printing usage
- stable-2.4 | runtime: Stop getting OOM events from agent for "ttrpc closed" error
- kata-deploy: fix version bump from -rc to stable
- stable-2.4: release: Include all the rust vendored code into the vendored tarball
- stable-2.4 | tools: release: Do not consider release candidates as stable releases
- agent: Signal the whole process group
- stable-2.4 | docs: Update k8s documentation
- backport main commits to stable 2.4
- stable-2.4: Bump QEMU to 6.2 (bringing then SGX support in)
- runtime: Properly handle ESRCH error when signaling container
- stable-2.4 | versions: Upgrade to Cloud Hypervisor v22.1
f2319d69 release: Adapt kata-deploy for 2.4.0
cae48e9c agent: fix container stop error with signal SIGRTMIN+3
342aa95c kata-monitor: fix duplicated output when printing usage
9f75e226 runtime: add logs around sandbox monitor
363fbed8 runtime: stop getting OOM events when ttrpc: closed error
f840de5a workflows,release: Ship *all* the rust vendored code
952cea5f tools: Add a generate_vendor.sh script
cc965fa0 kata-deploy: fix version bump from -rc to stable
f41cc184 tools: release: Do not consider release candidates as stable releases
e059b50f runtime: Add more debug logs for container io stream copy
71ce6f53 agent: Kill the all the container processes of the same cgroup
30fc2c86 docs: Update k8s documentation
24028969 virtcontainers: Run mock hook from build tree rather than system bin dir
4e54aa5a doc: fix filename typo
d815393c manager: Add options to change self test behaviour
4111e1a3 manager: Add option to enable component debug
2918be18 manager: Create containerd link
6b31b068 kernel: fix cve-2022-0847
5589b246 doc: update Intel SGX use cases document
1da88dca tools: update QEMU to 6.2
3e2f9223 runtime: Properly handle ESRCH error when signaling container
4c21cb3e versions: Upgrade to Cloud Hypervisor v22.1
Signed-off-by: Peng Tao <bergwolf@hyper.sh>
kata-deploy files must be adapted to a new release. The cases where it
happens are when the release goes from -> to:
* main -> stable:
* kata-deploy-stable / kata-cleanup-stable: are removed
* stable -> stable:
* kata-deploy / kata-cleanup: bump the release to the new one.
There are no changes when doing an alpha release, as the files on the
"main" branch always point to the "latest" and "stable" tags.
Signed-off-by: Peng Tao <bergwolf@hyper.sh>
The nix::sys::signal::Signal package api cannot deal with SIGRTMIN+3,
directly use libc function to send the signal.
Fixes: #3990
Signed-off-by: Wang Xingxing <stellarwxx@163.com>
(cherry picked from commit 0d765bd082)
Signed-off-by: Wang Xingxing <stellarwxx@163.com>
(default: "/run/containerd/containerd.sock") is duplicated when
printing kata-monitor usage:
[root@kubernetes ~]# kata-monitor --help
Usage of kata-monitor:
-listen-address string
The address to listen on for HTTP requests. (default ":8090")
-log-level string
Log level of logrus(trace/debug/info/warn/error/fatal/panic). (default "info")
-runtime-endpoint string
Endpoint of CRI container runtime service. (default: "/run/containerd/containerd.sock") (default "/run/containerd/containerd.sock")
the golang flag package takes care of adding the defaults when printing
usage. Remove the explicit print of the value so that it would not be
printed on screen twice.
Fixes: #3998
Signed-off-by: Francesco Giudici <fgiudici@redhat.com>
(cherry picked from commit a63bbf9793)
getOOMEvents is a long-waiting call, it will retry when failed.
For cases of agent shutdown, the retry should stop.
When the agent hasn't detected agent has died, we can also check
whether the error is "ttrpc: closed".
Fixes: #3815
Signed-off-by: bin <bin@hyper.sh>
