If the resource handle has data from a structured parameter model, then we need
to pass that to the DRA driver kubelet plugin. Because Kubernetes uses
gogo/protobuf, we cannot use "optional" for that new optional field and have to
resort to "repeated" with a single repetition if present.
This is a new, backwards-compatible field.
That extending the resource.k8s.io changes the checksum of a kubelet checkpoint
is unfortunate. Updating the test cases is a stop-gap measure, the actual
solution will have to be something else before beta.
Like the current device plugin interface, a DRA driver using this model
announces a list of resource instances. In contrast to device plugins, this
list is made available to the scheduler together with attributes that can be
used to select suitable instances when they are not all alike.
Because this is the first structured parameter model, some checks that
previously were not possible, in particular "is one structured parameter field
set", now gets enabled. Adding another structured parameter model will be
similar.
The applyconfigs code generator assumes that all types in an API are defined in
a single package. If it wasn't for that, it would be possible to place the
"named resources" types in separate packages, which makes their names in the Go
code more natural and provides an indication of their stability level because
the package name could include a version.
When a claim uses structured parameters, as indicated by the resource class
flag, the scheduler is responsible for allocating it. To do this it needs to
gather information about available node resources by watching
NodeResourceSlices and then match the in-tree claim parameters against those
resources.
NodeResourceSlice will be used by kubelet to publish resource information on
behalf of DRA drivers on the node. NodeName and DriverName in
NodeResourceSlice must be immutable. This simplifies tracking the different
objects because what they are for cannot change after creation.
The new field in ResourceClass tells scheduler and autoscaler that they are
expected to handle allocation.
ResourceClaimParameters and ResourceClassParameters are new types for telling
in-tree components how to handle claims.
* support for the managed-by label in Job
* Use managedBy field instead of managed-by label
* Additional review remarks
* Review remarks 2
* review remarks 3
* Skip cleanup of finalizers for job with custom managedBy
* Drop the performance optimization
* imrpove logs
Since the migration to the external cloud providers, the node
controller in the cloud controller manager is responsible of
initializing the nodes.
There is a strong assumption across the ecosystem that the nodes has
set the node.spec.providerID value, however, the node controller does
not check if this value is set during the initialization of the node,
and if there are some failures on the cloud provider API calls, the
node can be untainted without the value and never reconciled.
In addition, it seems that is possible for some cloud provider to not
implement the providerID value, though is not likely this is going to
happen, but for backward compatibility purposes we should allow this case.
The node controller will require the providerID to untain the Nodes,
except when the cloud provider does not use InstancesV2 and does implement it.
ProviderID is inmutable once set, so that value has preferences,
otherwise InstancesV2 is preferred over Instances.
Change-Id: Ic41cf7ebcca1ff0fbd8daafc036166f19fc37251
Signed-off-by: Antonio Ojea <aojea@google.com>
InitialEventsAnnotationKey the name of the key
under which an annotation marking the end of
a watchlist stream is stored.
The annotation is added to a "Bookmark" event.
The const will be immediately used in client-go and apiserver packages.
