Not sure how and when it got broken, but we're looking for mounts for
the blockdevice (like `/dev/vda`), while the actual mount info contains
the partition device (like `/dev/vda6`).
Signed-off-by: Andrey Smirnov <andrey.smirnov@talos-systems.com>
Having polymorphic (spec type depends on ID) resources is not a good
idea, and it's not compatible with protobuf encoding.
Introduce new resources for each polymorphic sub-spec using new Go 1.18
generic typed.Resource to reduce the boilerplate code.
(Still needs proper deepcopy-gen, but I'm skipping it for now, as
K8sControlPlane had also broken deep copy).
Signed-off-by: Andrey Smirnov <andrey.smirnov@talos-systems.com>
With update of the client library to 3.5.3, etcd library started using
the logger, so using `nil` isn't fine anymore.
Signed-off-by: Andrey Smirnov <andrey.smirnov@talos-systems.com>
Add a note on how machine configuration can be retrieved
from the node, after e.g. interactive setup.
Signed-off-by: Tim Jones <tim.jones@siderolabs.com>
Many users have been using the VIP functionality to configure
endpoints in Talos config. Documentation to clarify the possible
issues with that option and that it should be avoided.
Signed-off-by: Tim Jones <tim.jones@siderolabs.com>
This "fixes" the message like:
```
xfs filesystem being mounted at /var supports timestamps until 2038 (0x7fffffff)
```
We should support Talos beyond 2038, even if we switch to a different
filesystem type by 2038 :)
Signed-off-by: Andrey Smirnov <andrey.smirnov@talos-systems.com>
Fixes a typo in the Extension Services document alias
which serves as the redirect from the old location.
Signed-off-by: Tim Jones <tim.jones@siderolabs.com>
Signed-off-by: Andrey Smirnov <andrey.smirnov@talos-systems.com>
Enable the Rpi4 PoE hat fan control by pulling in the overlay
compatible with the upstream kernel driver.
Ref: https://github.com/siderolabs/pkgs/pull/450
Signed-off-by: Noel Georgi <git@frezbo.dev>
Dry run prints out config diff, selected application mode without
changing the configuration.
Signed-off-by: Artem Chernyshev <artem.chernyshev@talos-systems.com>
Containerd CRI plugin was merged into the main repo, but we were using
old import path, so our constants coming from the module were outdated.
This fixes the image version for the pause container.
Signed-off-by: Andrey Smirnov <andrey.smirnov@talos-systems.com>
See https://github.com/etcd-io/etcd/releases/tag/v3.5.3
This release should contain a fix for data consistency issue when etcd
is killed under high load.
Signed-off-by: Andrey Smirnov <andrey.smirnov@talos-systems.com>
Make improvements to help documentation discoverability and categorization.
Ensure all content pages have a description.
Ensure all link are replaced with Hugo shortcode.
Ensure all moved pages have an alias so redirects work.
Signed-off-by: Tim Jones <tim.jones@siderolabs.com>
Bump tools and pkgs to get kernel 5.15.33
5.15.33 has a bunch of fixes for some CVE's,
it was too hard to track those and reference
Signed-off-by: Noel Georgi <git@frezbo.dev>
Increase go.mod version from 1.17 to 1.18 in all projects. Update Makefile
to use latest tooling. Fix golangci by disable nolintlint for now.
Signed-off-by: Dmitriy Matrenichev <dmitry.matrenichev@siderolabs.com>
`talos.dev` redirects to `www.talos.dev` so this fix to
the sitemap removes the redirect when scraping.
Signed-off-by: Tim Jones <tim.jones@siderolabs.com>
Use the last `:` in the image reference.
Handle the case when no version was discovered.
See https://github.com/siderolabs/theila/issues/138
Signed-off-by: Andrey Smirnov <andrey.smirnov@talos-systems.com>
This PR enables the little arrow that shows there's stuff underneath a
header on the sidebar.
Signed-off-by: Spencer Smith <spencer.smith@talos-systems.com>
People are searching talos docs for cluserctl, so mentioning in this page to point them towards Sidero Metal.
Signed-off-by: Steve Francis <steve.francis@talos-systems.com>
Signed-off-by: Andrey Smirnov <andrey.smirnov@talos-systems.com>
This change fixes an issue where a code fence unintentionally included
the entire rest of the page.
Signed-off-by: William Ashton <William@AshtonFam.org>
Signed-off-by: Andrey Smirnov <andrey.smirnov@talos-systems.com>
This can be used when probing filesystems with config data in the
`metal-iso` case or in the `nocloud` platform.
Small disk images by default are created with small bits FAT
filesystems.
See https://github.com/siderolabs/go-blockdevice/pull/59
Signed-off-by: Andrey Smirnov <andrey.smirnov@talos-systems.com>
This PR overrides the sitemap to only include, for example, v1.0 results
if that's our specified latest version. This should fix Google's
indexing of our docs to only include the latest results.
Signed-off-by: Spencer Smith <spencer.smith@talos-systems.com>
Fix a typo in the QEMU VM docs.
Signed-off-by: Andrei Dobre <andreidobre.web@gmail.com>
Signed-off-by: Andrey Smirnov <andrey.smirnov@talos-systems.com>
The main goal to avoid cache invalidation on `tag` changes: by pushing
the tag down the build after the generate step, avoid invalidating build
cache on each tag change (`-dirty`, rebasing, etc.)
Signed-off-by: Andrey Smirnov <andrey.smirnov@talos-systems.com>
Docker by default disable IPv6 completely in the containers which breaks
SideroLink on Docker-based clusters, as SideroLink is using IPv6
addresses for the Wiregurard tunnel.
This change might break `talosctl cluster create` on host systems which
have IPv6 disabled completely, so provide a flag to revert this
behavior.
Signed-off-by: Andrey Smirnov <andrey.smirnov@talos-systems.com>
This PR is an omnibus edition of Steve's PRs. It includes three small
things to the docs:
- A change to config.toml to avoid redirects a bit better
- Updates to digital rebar doc that fixes a few links and typos.
- Updates for typos and wording of storage.md
Signed-off-by: Spencer Smith <spencer.smith@talos-systems.com>
This showed up recently frequently in integration-provision tests
(might be related to Kubernetes upgrade), but anyways errors should be
retried.
Refactored the function to extract the retryable part.
Signed-off-by: Andrey Smirnov <andrey.smirnov@talos-systems.com>
In the case of a node being reset, using kexec greatly
speeds up the process. However, in the event the boot
partition is wiped, a full reboot is required.
Closes#4670
Signed-off-by: Tim Jones <tim.jones@siderolabs.com>
