Changes `gen config` to output `controlplane` and `join` machine config
types only. Users can manually set the `type` to `init` if they need to.
Signed-off-by: Andrew Rynhard <andrew@rynhard.io>
The structure of the controllers is really similar to addresses and
routes:
* `LinkSpec` resource describes desired link state
* `LinkConfig` controller generates `LinkSpecs` based on machine
configuration and kernel cmdline
* `LinkMerge` controller merges multiple configuration sources into a
single `LinkSpec` paying attention to the config layer priority
* `LinkSpec` controller applies the specs to the kernel state
Controller `LinkStatus` (which was implemented before) watches the
kernel state and publishes current link status.
Signed-off-by: Andrey Smirnov <smirnov.andrey@gmail.com>
Fixes: https://github.com/talos-systems/talos/issues/3556
Introduce `coredns.disabled` flag to the machine config that allows
turning off coredns deployment during cluster bootstrap.
Signed-off-by: Artem Chernyshev <artem.0xD2@gmail.com>
* without `-print0` / `-0`, paths with spaces were skipped sometimes;
* `--stdin-filename` flag incorrectly consumed the first file.
Signed-off-by: Alexey Palazhchenko <alexey.palazhchenko@gmail.com>
This PR adds some documentation on how to use mayastor with talos and
links the video showing it in action.
Signed-off-by: Spencer Smith <robertspencersmith@gmail.com>
The problem is that some patches can't be applied to join config, as
some nodes don't even exist in the config, for example
`/cluster/apiServer` node, and applying such patches doesn't make any
sense.
Signed-off-by: Andrey Smirnov <smirnov.andrey@gmail.com>
This seems to be unused completely, and they look scary enough at the
same time.
For better readability and to avoid any pitfalls, better to remove them.
Signed-off-by: Andrey Smirnov <smirnov.andrey@gmail.com>
Describe full procedure from backups to disaster recovery.
Signed-off-by: Andrey Smirnov <smirnov.andrey@gmail.com>
Co-authored-by: Spencer Smith <rsmitty@users.noreply.github.com>
Command implements two modes:
* `fast`: conformance suite is run at maximum speed
* `certified`: conformance suite is run in serial mode, results
are capture to produce artifacts ready for CNCF submission process
Signed-off-by: Andrey Smirnov <smirnov.andrey@gmail.com>
Sometimes `talosctl etcd snapshot` might not be available, for example
when etcd is not healthy. In that case it's possible to copy raw etcd
data directory with `talosctl cp /var/lib/etcd .` and use
`member/snap/db` to recover the cluster. But such copy won't pass
integrity checks, so they should be disabled explicitly.
Signed-off-by: Andrey Smirnov <smirnov.andrey@gmail.com>
For config patching, it's beneficial to have some fields to be present
in the config, even with the default (empty) value. At the same time not
all fields should be present in all versions of the config, e.g. some
config value don't apply to worker node configuration.
Empty value and nil value are treated equal by `yaml` library, but Talos
encoder can be made more smart to still output empty (non-nil) value to
the config, while skipping completely nil fields.
This PR implements that via new `talos:"omitonlyifnil"` tag and plus
moves docs for such fields into comments under the value.
GC'ed pod checkpointer config, so it doesn't get generated even as
comments (it was empty by default even for 0.8, so this just removes
comments about it).
Signed-off-by: Andrey Smirnov <smirnov.andrey@gmail.com>
Inline manifests work exactly same way as extra manifests, but the
manifest itself can be stored in the config body.
Example config patch:
```
--config-patch '[{"op": "replace", "path": "/cluster/inlineManifests", "value": [{"name": "foo", "contents": "apiVersion: v1\nkind: Namespace\nmetadata:\n name: ci\n"}]}]'
```
Signed-off-by: Andrey Smirnov <smirnov.andrey@gmail.com>
Fixes a typo in a documented command for going from an unencrypted
EPHEMERAL partition to an encrypted one.
Signed-off-by: Brandon McNama <brandonmcnama@outlook.com>
When Talos `controlplane` node is waiting for a bootstrap, `etcd`
contents can be recovered from a snapshot created with
`talosctl etcd snapshot` on a healthy cluster.
Bootstrap process goes same way as before, but the etcd data directory
is recovered from the snapshot.
This flow enables disaster recovery for the control plane: given that
periodic backups are available, destroy control plane nodes, re-create
them with the same config, and bootstrap one node with the saved
snapshot to recover etcd state at the time of the snapshot.
Signed-off-by: Andrey Smirnov <smirnov.andrey@gmail.com>
