Kubeconfig is merged into `~/.kube/config` with rename option
(existing configuration is never overwritten).
If endpoint was used, it is automatically put into the `kubeconfig`.
This should make OS X experience literally `talosctl cluster create`
followed by any `kubectl get ...`.
Signed-off-by: Andrey Smirnov <smirnov.andrey@gmail.com>
By default, build outside of Drone works the same and builds only amd64
version, loads images back into dockerd, etc.
If multiple platforms are used, multi-arch images are built which can't
be exported to docker or to `.tar` image, they're always pushed to the
registry (even for PR builds to our internal CI registry).
Artifacts as files (initramfs, kernel) now have `-arch` suffix:
`vmlinuz-amd64`, `initramfs-amd64.xz`. "Magic" script normalizes output
paths depending on whether single platform or multiple platforms were
given.
VM provisioners accept magic `${ARCH}` in initramfs/kernel paths which
gets replaced by cluster architecture.
Signed-off-by: Andrey Smirnov <smirnov.andrey@gmail.com>
Our defaults were too low today, and for Docker they're purely limits,
not pre-allocations. For QEMU we need more memory by default, and QEMU
doesn't support fractional CPU resource anyways.
Disk is a sparse file, so it's not allocated unless being used.
Fixes#2561
Signed-off-by: Andrey Smirnov <smirnov.andrey@gmail.com>
As we're going to make API bootstrapping a preferred method of
bootstrapping Talos clusters, our defaults should reflect that.
Signed-off-by: Andrey Smirnov <smirnov.andrey@gmail.com>
Fixes#2363#2364#2370#2371
Several changes packed together:
* use compressed `vmlinuz` everywhere, firecracker provisioner
uncompresses it before first use, drop `vmlinux`
* handle reboots in qemu launcher to support reset API case, update
empty disk check to handle reset behavior (erasing partition table)
* make bootloader support default in provisioners, and flag to disable
that
* early support for target architecture for qemu provisioner
This should allow us to use `qemu` in CI/CD (not included into this PR):
integration test passes with qemu.
Signed-off-by: Andrey Smirnov <smirnov.andrey@gmail.com>
Starts and stops qemu VMs, has some initial configuration subset.
Sets up networking through CNI tools, sets up DHCP server which gives IP
addresses to nodes.
Signed-off-by: Artem Chernyshev <artem.0xD2@gmail.com>
When cluster fails to be bootstrapped or it fails the health check, it's
hard to find the root cause without the logs.
This change adds optional crashdump (it dumps firecracker logs or docker
logs) after provisioning failure. It's not enabled by default.
Signed-off-by: Andrey Smirnov <smirnov.andrey@gmail.com>
This adds new flug to `cluster create` to launch cluster with custom
CNI, `integration` pipeline gets a new step to run short test with
Cilium 1.8.0 CNI.
Signed-off-by: Andrey Smirnov <smirnov.andrey@gmail.com>
This PR brings in all changes necessary to deploy kubernetes 1.19.x.
It relies on an update to our bootkube-plugin project, as well as
implementation of some Image() functions for our various control plane
components, since they are all distinct images and not just hyperkube.
Signed-off-by: Spencer Smith <robertspencersmith@gmail.com>
This PR will update k8s to the latest 1.18 release and bump sonobuoy to
help resolve some e2e flakes. Also adds some retry logic around the
sonobuoy run.
Signed-off-by: Spencer Smith <robertspencersmith@gmail.com>
This PR will fix some unexpected user behavior where nameservers were
always getting written to 8.8.8.8,1.1.1.1 for the docker-based talos
clusters. This occurred even when updating the docker daemon's config.
This PR will make the docker provisioner respect the --nameserver flag
and allow that to be used to override the defaults.
Signed-off-by: Spencer Smith <robertspencersmith@gmail.com>
This adds the ability to bootstrap a cluster using the API.
The API simply starts the bootkube service.
Signed-off-by: Andrew Rynhard <andrew@andrewrynhard.com>
This PR will introduce a `-p/--exposed-ports` flag to talosctl. This
flag will allow us to enable port forwards on worker nodes only. This
will allow for ingresses on docker clusters so we can hopefully use
ingress for Arges initial bootstrapping. I modeled this after how KIND allows ingresses
[here](https://kind.sigs.k8s.io/docs/user/ingress/)
Signed-off-by: Spencer Smith <robertspencersmith@gmail.com>
This PR will pull in the latest release of k8s 1.18 so we can start
validating it through our test suite.
Signed-off-by: Spencer Smith <robertspencersmith@gmail.com>
It seems to be useful enough to be the default one and it prevents
simple mistakes while trying to access the cluster which is not ready
yet.
Signed-off-by: Andrey Smirnov <smirnov.andrey@gmail.com>
This is a rename of the osctl binary. We decided that talosctl is a
better name for the Talos CLI. This does not break any APIs, but does
make older documentation only accurate for previous versions of Talos.
Signed-off-by: Andrew Rynhard <andrew@andrewrynhard.com>
