add option to eneable only pod cgroup (SandboxCgroupOnly)
Depends-on: github.com/kata-containers/tests#1824
Fixes: #1879
Signed-off-by: Jose Carlos Venegas Munoz <jose.carlos.venegas.munoz@intel.com>
Since virtio-fs is under active development, more
options will be added increasingly. To avaoid frequent
change on runtime side to handle option changes, use
one mingled arg to ease testing new option/feature of
virtiofsd.
See `virtiofsd -h` for more option details.
Fixes: #1999
Signed-off-by: Eric Ren <renzhen@linux.alibaba.com>
The list of kernel modules can be passed to the runtime through the
configuration file or using OCI annotations. In both cases, a list paramentes
can be specified for each module.
fixes#1925
Signed-off-by: Julio Montes <julio.montes@intel.com>
Add jailer support to configuration files.
Also enable jailer by default in Kata containers.
Signed-off-by: Manohar Castelino <manohar.r.castelino@intel.com>
This patch covers the following aspects,
1. Add ACRN as a supported hypervisor for amd64 architecture.
2. Build and install configuration file for ACRN hypervisor.
v1->v2:
1. Deleted autogenerated configuration-acrn.toml.
2. Trimmed configuration options not used by ACRN.
Fixes: #1778
Signed-off-by: Vijay Dhanraj <vijay.dhanraj@intel.com>
proxy will never be use with the Firecracker VMM. Keeping this header
will result in runtime failures, since the configuration will be parsed
on the path searched for.
Since vsock will always be used, remove the proxy section.
Fixes: #1761
Signed-off-by: Eric Ernst <eric.ernst@intel.com>
A file based memory backend mapped to the host, fot eg: '/dev/shm' will
be used by virtio-fs for performance reasons. This change is a generic
implementation of that for kata. This will be enabled default for
virtio-fs negating the need to enable hugepages in that scenario. This
option can be used without virtio-fs by setting 'file_mem_backend' to
the location in the configuration file. Default value is an empty
string.
Fixes: #1656
Signed-off-by: Ganesh Maharaj Mahalingam <ganesh.mahalingam@intel.com>
nemu needs to be configured with:
`machine_type = "virt"` by default.
In addition, this commit removes
`machine_accelerators="virt"` which was added instead
of `machine_type` in a previous commit.
Fixes: #1707.
Signed-off-by: Salvador Fuentes <salvador.fuentes@intel.com>
- For initial offering of virtiofsd, hugepages are required
- use the qemu hypervisor type for configuration template
- decrease virtiofs cache size from 8192 to 1024, to better support
running with virtio-fs on smaller machines while hugepages are required.
- For initial experimental release, utilize virtio-serial instead of
vsock
Fixes: #1662
Signed-off-by: Eric Ernst <eric.ernst@intel.com>
NEMU requires a unique runtime configuration. Add support for utilizing
a configuration-nemu.toml
Fixes: #1647
Signed-off-by: Eric Ernst <eric.ernst@intel.com>
Several cache modes are supported by virtio-fs. They affect the
performance and consistency characteristics of the file system.
For the time being cache="none" is recommended, but the other modes can
be experimented with.
Signed-off-by: Dr. David Alan Gilbert <dgilbert@redhat.com>
Add VirtioFSCacheSize aka virtio_fs_cache_size option
to set the size (in MiB) of the DAX cache.
Signed-off-by: Dr. David Alan Gilbert <dgilbert@redhat.com>
Add a config option to select between virtio-9p and virtiofs. This
option currently has no effect and will be used in a later patch.
Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>
Add configuration options to support the various Kata agent tracing
modes and types. See the comments in the built configuration files for
details:
- `cli/config/configuration-fc.toml`
- `cli/config/configuration-qemu.toml`
Fixes#1369.
Signed-off-by: James O. D. Hunt <james.o.hunt@intel.com>
Previously, the agent behaviour was controlled entirely using the
`kernel_params=` config option. This mechanism suffers from a subtle
problem - the runtime is not aware of how the agent will behave.
From now on, all significant agent options will be controlled from the
agent section in the configuration file. This allows the runtime to be
more aware of -- and in control of -- such agent settings. It would also
allow the underlying kernel CLI options to be modified in the future if
required.
This PR adds the only useful agent option as an explicit option by
adding an `enable_debug=true` option to the Kata agent section in
`configuration.toml`. This allows controlling agent debug to be handled
in the same manner as the other debug options.
This change is somewhat foundational: it permits the agent to be handled
consistently with other config file sections which is useful, but
arguably not essential (the old way worked). However, the new way of
handling agent options will be essential when introducing agent tracing
control as the runtime must be aware of the agent trace mode to allow
the runtime to modify its behaviour accordingly.
Signed-off-by: James O. D. Hunt <james.o.hunt@intel.com>
Fixes#1226
Add new flag "experimental" for supporting underworking features.
Some features are under developing which are not ready for release,
there're also some features which will break compatibility which is not
suitable to be merged into a kata minor release(x version in x.y.z)
For getting these features above merged earlier for more testing, we can
mark them as "experimental" features, and move them to formal features
when they are ready.
Signed-off-by: Wei Zhang <zhangwei555@huawei.com>
VMCache is a new function that creates VMs as caches before using it.
It helps speed up new container creation.
The function consists of a server and some clients communicating
through Unix socket. The protocol is gRPC in protocols/cache/cache.proto.
The VMCache server will create some VMs and cache them by factory cache.
It will convert the VM to gRPC format and transport it when gets
requestion from clients.
Factory grpccache is the VMCache client. It will request gRPC format
VM and convert it back to a VM. If VMCache function is enabled,
kata-runtime will request VM from factory grpccache when it creates
a new sandbox.
VMCache has two options.
vm_cache_number specifies the number of caches of VMCache:
unspecified or == 0 --> VMCache is disabled
> 0 --> will be set to the specified number
vm_cache_endpoint specifies the address of the Unix socket.
This commit just includes the core and the client of VMCache.
Currently, VM cache still cannot work with VM templating and vsock.
And just support qemu.
Fixes: #52
Signed-off-by: Hui Zhu <teawater@hyper.sh>
Support building with `make HYPERVISOR=firecracker`.
Notes:
- Firecracker is currently only supported on x86_64, hence the
architecture check.
- The template file `cli/config/configuration-fc.toml.in` is based
on `cli/config/configuration.toml.in`.
Fixes#1105.
Signed-off-by: James O. D. Hunt <james.o.hunt@intel.com>
Improve the build to:
- Encapsulate the qemu configuration.
- Validate the chosen hypervisor.
- Display hypervisor details for `make help`.
This lays the groundwork to build configurations for multiple
hypervisors.
Notes:
- Variables that are hypervisor specific have now been suffixed with `_QEMU`.
- `make install` now installs `configuration-qemu.toml`.
To maintain compatibility with existing installations, a symbolic link
to this file called `configuration.toml` is created.
Signed-off-by: James O. D. Hunt <james.o.hunt@intel.com>
Pass Seccomp profile to the agent only if
the configuration.toml allows it to be passed
and the agent/image is seccomp capable.
Fixes: #688
Signed-off-by: Nitesh Konkar niteshkonkar@in.ibm.com
This value will be plused to max memory of hypervisor.
It is the memory address space for the NVDIMM devie.
If set block storage driver (block_device_driver) to "nvdimm",
should set memory_offset to the size of block device.
Signed-off-by: Hui Zhu <teawater@hyper.sh>
Set block_device_driver to "nvdimm" will make the hypervisor use
the block device as NVDIMM disk.
Fixes: #1032
Signed-off-by: Hui Zhu <teawater@hyper.sh>
If VM factory templating is enabled (`enable_template=true`), error if
the configured image is not an `initrd=` one.
Also add a note to the config file explaining that a normal image cannot
be used - only initrd images are supported.
Fixes#948.
Signed-off-by: James O. D. Hunt <james.o.hunt@intel.com>
Add configuration to decide the amount of slots that will be used in a VM
- This will limit the amount of times that memory can be hotplugged.
- Use memory slots provided by user.
- tests: aling struct
cli: kata-env: Add memory slots info.
- Show the slots to be added to the VM.
```diff
[Hypervisor]
MachineType = "pc"
Version = "QEMU ..."
Path = "/opt/kata/bin/qemu-system-x86_64"
BlockDeviceDriver = "virtio-scsi"
Msize9p = 8192
+ MemorySlots = 10
Debug = false
UseVSock = false
```
Fixes: #751
Signed-off-by: Jose Carlos Venegas Munoz <jose.carlos.venegas.munoz@intel.com>
In order to choose if the network monitor should be used or not, this
patch makes it configurable from the configuration.toml file.
Signed-off-by: Sebastien Boeuf <sebastien.boeuf@intel.com>
Add `disable_vhost_net` option to enable or disable the use of
vhost_net. Vhost_net can improve network performance.
Signed-off-by: Ruidong Cao <caoruidong@huawei.com>
We need this configuration due to a limitation in seabios
firmware in handling hotplug for PCI devices with large BARS.
Long term, this needs to be fixed in the firmware.
Fixes#594
Signed-off-by: Archana Shinde <archana.m.shinde@intel.com>
Add initial support for opentracing by using the `jaeger` package.
Since opentracing uses the `context` package, add a `context.Context`
as the first parameter to all the functions that we might want to
trace. Trace "spans" (trace points) are then added by extracting the
trace details from the specified context parameter.
Notes:
- Although the tracer is created in `main()`, the "root span"
(aka the first trace point) is not added until `beforeSubcommands()`.
This is by design and is a compromise: by delaying the creation of the
root span, the spans become much more readable since using the web-based
JaegerUI, you will see traces like this:
```
kata-runtime: kata-runtime create
------------ -------------------
^ ^
| |
Trace name First span name
(which clearly shows the CLI command that was run)
```
Creating the span earlier means it is necessary to expand 'n' spans in
the UI before you get to see the name of the CLI command that was run.
In adding support, this became very tedious, hence my design decision to
defer the creation of the root span until after signal handling has been
setup and after CLI options have been parsed, but still very early in
the code path.
- At this stage, the tracing stops at the `virtcontainers` call
boundary.
- Tracing is "always on" as there doesn't appear to be a way to toggle
it. However, its resolves to a "nop" unless the tracer can talk to a
jaeger agent.
Note that this commit required a bit of rework to `beforeSubcommands()`
to reduce the cyclomatic complexity.
Fixes#557.
Signed-off-by: James O. D. Hunt <james.o.hunt@intel.com>
Add `use_vsock` option to enable or disable the use of vsocks
for communication between host and guest.
Signed-off-by: Jose Carlos Venegas Munoz <jose.carlos.venegas.munoz@intel.com>
Signed-off-by: Julio Montes <julio.montes@intel.com>
Add enable_template option to the config file.
When it is set, enable the vm template factory.
cache factory cannot be used by kata cli directly because
it requires a running daemon to maintain the cache VMs.
`kata-runtime factory init` would initialize the vm factory and
`kata-runtime factory destroy` would destroy the vm factory.
When configured, a vm factory is loaded before creating new sandboxes.
Signed-off-by: Peng Tao <bergwolf@gmail.com>
There is a relation between the maximum number of vCPUs and the
memory footprint, if QEMU maxcpus option and kernel nr_cpus
cmdline argument are big, then memory footprint is big, this
issue only occurs if CPU hotplug support is enabled in the kernel,
might be because of kernel needs to allocate resources to watch all
sockets waiting for a CPU to be connected (ACPI event).
For example
```
+---------------+-------------------------+
| | Memory Footprint (KB) |
+---------------+-------------------------+
| NR_CPUS=240 | 186501 |
+---------------+-------------------------+
| NR_CPUS=8 | 110684 |
+---------------+-------------------------+
```
In order to do not affect CPU hotplug and allow to users to have containers
with the same number of physical CPUs, this patch tries to mitigate the
big memory footprint by using the actual number of physical CPUs as the
maximum number of vCPUs for each container if `default_maxvcpus` is <= 0 in
the runtime configuration file, otherwise `default_maxvcpus` is used as the
maximum number of vCPUs.
Before this patch a container with 256MB of RAM
```
total used free shared buff/cache available
Mem: 195M 40M 113M 26M 41M 112M
Swap: 0B 0B 0B
```
With this patch
```
total used free shared buff/cache available
Mem: 236M 11M 188M 26M 36M 186M
Swap: 0B 0B 0B
```
fixes#295
Signed-off-by: Julio Montes <julio.montes@intel.com>
The config file is in TOML format which supports comments, so add the
license header to it.
Fixes#234.
Signed-off-by: James O. D. Hunt <james.o.hunt@intel.com>
As agreed in [the kata containers API
design](https://github.com/kata-containers/documentation/blob/master/design/kata-api-design.md),
we need to rename pod notion to sandbox. The patch is a bit big but the
actual change is done through the script:
```
sed -i -e 's/pod/sandbox/g' -e 's/Pod/Sandbox/g' -e 's/POD/SB/g'
```
The only expections are `pod_sandbox` and `pod_container` annotations,
since we already pushed them to cri shims, we have to use them unchanged.
Fixes: #199
Signed-off-by: Peng Tao <bergwolf@gmail.com>
Add option to configure if IO needs to be in a separate IO thread.
Add tests to verify option is correctly parsed. The default value
is set to false for now.
This should be considered to be enabled by default in the future.
Fixes#132
Signed-off-by: Archana Shinde <archana.m.shinde@intel.com>
