blob: 28becd3d4f39c7d77ffae8d12a1a2468bf73604a [file] [log] [blame]
// Copyright 2020 The Monogon Project Authors.
//
// SPDX-License-Identifier: Apache-2.0
//
// Licensed under the Apache License, Version 2.0 (the "License");
// you may not use this file except in compliance with the License.
// You may obtain a copy of the License at
//
// http://www.apache.org/licenses/LICENSE-2.0
//
// Unless required by applicable law or agreed to in writing, software
// distributed under the License is distributed on an "AS IS" BASIS,
// WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
// See the License for the specific language governing permissions and
// limitations under the License.
syntax = "proto3";
package metropolis.proto.common;
option go_package = "source.monogon.dev/metropolis/proto/common";
import "google/protobuf/timestamp.proto";
// NodeRoles are the possible roles that a Metropolis Node should run within the
// cluster. These are configured by the cluster and can be retrieved through the
// Curator.
//
// Fields contained within each individual are publicly available, so while they
// can be used to carry required data to start up services for a given role,
// this must not be confidential/private data.
message NodeRoles {
message KubernetesController {
}
message KubernetesWorker {
}
message ConsensusMember {
// ca_certificate is a DER-encoded x509 certificate of the etcd
// cluster's CA. The member must use this certificate to verify the
// identity of the cluster it's connecting to.
bytes ca_certificate = 1;
// pper_certificate is a DER-encoded x509 certificate of this node's
// etcd peer listener. The member must serve member traffic using this
// certificate. The private key corresponding to this certificate is
// the same as the node's primary private keypair.
bytes peer_certificate = 2;
// initial_crl is a certificate revocation list that the etcd member
// should be started with. After startup, the member will maintain its
// own CRL by updating it from its primary storage location, and etcd
// value.
//
// TODO(q3k): don't pass this here, instead pass this over an etcd
// watcher and curator.Watch.
bytes initial_crl = 3;
message Peer {
string Name = 1;
string URL = 2;
}
repeated Peer peers = 4;
}
KubernetesWorker kubernetes_worker = 1;
ConsensusMember consensus_member = 2;
KubernetesController kubernetes_controller = 3;
}
// NodeState is the state of a Metropolis node from the point of view of the
// cluster it is a part of (or intending to be a part of).
enum NodeState {
NODE_STATE_INVALID = 0;
// NEW: the node has established a first contact with the cluster and
// intends to register into it. The node's identity has not been verified
// and no hardware attestation of the new node was performed.
// The node has generated a CUK/NUK and set up storage encrypted with the
// combination of both keys.
// The node has generated a private/public keypair, and that keypair has
// been used to contact the already running Cluster.
NODE_STATE_NEW = 1;
// STANDBY: the node has successfully passed identity and hardware
// attestation checks as defined by the cluster policy. The node still isn't
// part of the cluster, as it itself might perform checks against the
// running Cluster.
NODE_STATE_STANDBY = 2;
// UP: the node has passed all preconditions for STANDBY and has also
// performed a commit into the cluster by exchanging its CUK for a
// certificate issued by the cluster.
// The node is now ready to serve, and its certificate can be used to
// authenticate its identity cryptographically.
NODE_STATE_UP = 3;
// DISOWNED: the node has been rejected or decommissioned by the cluster.
// Any further contact from the node to the cluster will be rejected.
NODE_STATE_DISOWNED = 4;
};
// ClusterState is the state of the cluster from the point of view of a node.
// Different subsystems can watch this state and depend on it for behaviour
// (eg. start serving when HOME, maybe self-fence on SPLIT, etc.).
enum ClusterState {
CLUSTER_STATE_INVALID = 0;
// UNKNOWN: the node has not yet determined the existence of a cluster it
// should join or start. This is a transient, initial state that should only
// manifest during boot.
CLUSTER_STATE_UNKNOWN = 1;
// FOREIGN: the node is attempting to register into an already existing
// cluster with which it managed to make preliminary contact, but which the
// cluster has not yet fully productionized (eg. the node is still being
// hardware attested, or the operator needs to confirm the
// registration of this node).
CLUSTER_STATE_FOREIGN = 2;
// TRUSTED: the node is attempting to register into an already registered
// cluster, and has been trusted by it. The node is now attempting to fully
// commit to registering into the cluster.
CLUSTER_STATE_TRUSTED = 3;
// HOME: the node is part of this cluster. This is the bulk of time in which
// this node will spend its time.
CLUSTER_STATE_HOME = 4;
// DISOWNING: the node has been disowned (ie., removed) by the cluster, and
// that it will not be ever part of any cluster again, and that it will be
// decommissioned by the operator.
CLUSTER_STATE_DISOWNING = 5;
// SPLIT:the node would usually be Home in a cluster, but has been split
// from the consensus of the cluster. This can happen for nodes running
// consensus when consensus is lost (eg. when there is no quorum or this
// node has been netsplit), and for other nodes if they have lost network
// connectivity to the consensus nodes. Clients should make their own
// decision what action to perform in this state, depending on the level of
// consistency required and whether it makes sense for the node to fence its
// services off.
CLUSTER_STATE_SPLIT = 6;
}
// NodeStatus contains all fields self-reported by nodes. This data is
// inherently less trusted than other data available about a node, as it can be
// updated to any value by each node individually, including compromised nodes.
message NodeStatus {
// external_address is the IP address that the node expects management,
// cluster and user traffic to arrive at (ie. the address on which it is
// listening for gRPC, and role-specific services like etcd and
// Kubernetes).
string external_address = 1;
// running_curator contains information about the curator service running
// on this node, or is nil if the service is not running.
message RunningCurator {
// port is the TCP port on which the curator is listening.
int32 port = 1;
}
RunningCurator running_curator = 3;
// timestamp is an epoch number associated with the last status update.
// It's set with a nanosecond granularity.
google.protobuf.Timestamp timestamp = 2;
}
// The Cluster Directory is information about the network addressing of nodes
// in a cluster. It is a serialized snapshot of some of the state within the
// etcd cluster, and can be used by external processes (like a node Registering
// into the cluster) to know how to reach this cluster over the network. It can
// be thought of as a phonebook, or a static name/address configuration that
// could live in /etc/hosts.
//
// The directory explicitly doesn't carry any information about the cluster's
// identity or security - these should be configured and checked by higher
// level configuration and processes. The directory can be stored and
// transmitted in cleartext and without an integrity checks (like saved to the
// EFI system partition across reboots) and any malicious change to it will
// cause no more than a denial of service against the consumer of this
// directory. This is because all nodes contacted must present a valid cluster
// identity/certificate before they are trusted by the consumers of this
// directory.
message ClusterDirectory {
message Node {
bytes public_key = 1;
message Address {
string host = 1;
};
repeated Address addresses = 2;
};
repeated Node nodes = 1;
}
// NodeClusterNetworking carries information about the cluster networking (ie.
// WireGuard mesh) connectivity of a node.
message NodeClusterNetworking {
message Prefix {
string cidr = 1;
}
// wireguard_pubkey is the base64-encoded public key used by the node.
string wireguard_pubkey = 1;
// prefixes are networking routes exported by the node to the cluster networking
// mesh, and are programmed by other nodes into their wireguard peer config.
repeated Prefix prefixes = 2;
}
// Severity level corresponding to //metropolis/pkg/logtree.Severity.
enum LeveledLogSeverity {
INVALID = 0;
INFO = 1;
WARNING = 2;
ERROR = 3;
FATAL = 4;
}
// Filter set when requesting logs for a given DN. This message is equivalent to
// the following GADT enum:
// data LogFilter = WithChildren
// | OnlyRaw
// | OnlyLeveled
// | LeveledWithMinimumSeverity(Severity)
//
// Multiple LogFilters can be chained/combined when requesting logs, as long as
// they do not conflict.
message LogFilter {
// Entries will be returned not only for the given DN, but all child DNs as
// well. For instance, if the requested DN is foo, entries logged to foo,
// foo.bar and foo.bar.baz will all be returned.
message WithChildren {
}
// Only raw logging entries will be returned. Conflicts with OnlyLeveled
// filters.
message OnlyRaw {
}
// Only leveled logging entries will be returned. Conflicts with OnlyRaw
// filters.
message OnlyLeveled {
}
// If leveled logs are returned, all entries at severity lower than `minimum`
// will be discarded.
message LeveledWithMinimumSeverity {
LeveledLogSeverity minimum = 1;
}
oneof filter {
WithChildren with_children = 1;
OnlyRaw only_raw = 3;
OnlyLeveled only_leveled = 4;
LeveledWithMinimumSeverity leveled_with_minimum_severity = 5;
}
}
// LogEntry corresponding to logtree.LogEntry in //metropolis/pkg/logtree.
message LogEntry {
// A leveled log entry emitted from a compatible system, eg. Metorpolis code
// or a klog-parsed line.
message Leveled {
repeated string lines = 1;
google.protobuf.Timestamp timestamp = 2;
LeveledLogSeverity severity = 3;
// Source of the error, expressed as file:line.
string location = 4;
}
// Raw log entry, captured from an external system without parting. Might
// contain some timestamp/level/origin information embedded in data. Data
// contained within should be treated as unsanitized external data.
message Raw {
string data = 1;
// Original length of line, set if data was truncated.
int64 original_length = 2;
}
// Origin DN.
string dn = 1;
oneof kind {
Leveled leveled = 2;
Raw raw = 3;
}
}
// ClusterConfiguration contains the entirety of the user-configurable behaviour
// of the cluster that is scoped to the entirety of the cluster (vs. per-node
// configuration, which is kept alongside Node).
//
// It can be set initially when a cluster is being bootstrapped (in
// NodeParamaters.ClusterBootstrap), and then can be partially managed by
// management calls to the curator.
message ClusterConfiguration {
// tpm_mode defines the TPM usage policy for cluster nodes. When nodes
// register into the cluster (and then join into it) they will report their
// TPM availability, and in return the cluster will respond whether they
// should use that TPM or not.
//
// If a node is instructed to use its TPM, it will use it to encrypt its part
// of the disk encryption key when saving it to the EFI system partition.
// That means that the node will only be able to re-join the cluster if its
// secure boot configuration doesn't change.
//
// If a node is instructed to not use its TPM, it will save its part of the
// disk encryption key straight onto the EFI system partition without any
// further encryption. It still needs to connect to a working cluster to
// retrieve the other part of the key. This means that the configuration is
// secure vs. offline disk decryption attempts, but not secure if an
// attacker can connect to a cluster and impersonate the node in order to
// retrieve the other part of its key.
enum TPMMode {
TPM_MODE_INVALID = 0;
// Nodes need to join with a TPM2.0 device and will be instructed to
// use it.
TPM_MODE_REQUIRED = 1;
// Nodes will be allowed to join regardless of TPM2.0 presence, and will
// be instructed to use it if they have one.
TPM_MODE_BEST_EFFORT = 2;
// Regardless of the node's local TPM presence it will be instructed to
// not use it.
TPM_MODE_DISABLED = 3;
}
TPMMode tpm_mode = 1;
// storage_security_policy defines which node storage security settings are
// accepted by the cluster. Nodes are informed of the cluster policy when
// registering into the cluster, alongside a cluster-recommended storage
// security setting. The node then reports its selected node storage setting
// during its Commit call which the cluster verifies against its policy.
enum StorageSecurityPolicy {
STORAGE_SECURITY_POLICY_INVALID = 0;
// The cluster accepts any storage security.
STORAGE_SECURITY_POLICY_PERMISSIVE = 1;
// The cluster accepts any storage security that offers encryption.
STORAGE_SECURITY_POLICY_NEEDS_ENCRYPTION = 2;
// The cluster accepts any storage security that offers encryption and
// authentication.
STORAGE_SECURITY_POLICY_NEEDS_ENCRYPTION_AND_AUTHENTICATION = 3;
// The cluster only accepts unencrypted and unauthenticated node storage.
STORAGE_SECURITY_POLICY_NEEDS_INSECURE = 4;
}
StorageSecurityPolicy storage_security_policy = 2;
}
// NodeTPMUsage describes whether a node has a TPM2.0 and if it is/should be
// actively used to seal secrets before saving them to its EFI system partition.
enum NodeTPMUsage {
NODE_TPM_INVALID = 0;
// This node has no TPM 2.0.
NODE_TPM_NOT_PRESENT = 1;
// This node has a TPM 2.0 but the cluster configuration mandates not using
// it.
NODE_TPM_PRESENT_BUT_UNUSED = 2;
// This node has a TPM 2.0 and it is being actively used.
NODE_TPM_PRESENT_AND_USED = 3;
}
// NodeStorageSecurity describes how a node encrypts and/or authenticates its
// local storage. In other words, it's a configuration setting for disk
// encryption (ie. via dm-crypt) and disk integrity (ie. via dm-integrity) of
// the Metropolis data partition.
enum NodeStorageSecurity {
NODE_STORAGE_SECURITY_INVALID = 0;
// The node has unencrypted and unauthenticated disk storage. Its data
// partition is a plain XFS partition, and the node's credentials are stored
// on it directly.
NODE_STORAGE_SECURITY_INSECURE = 1;
// The node has encrypted but unauthenticated disk storage. Its data
// partition is an XFS partition mounted through dm-crypt.
NODE_STORAGE_SECURITY_ENCRYPTED = 2;
// The node has encrypted and authenticated storage. Its data
// partition is an XFS partition mounted through dm-integrity and dm-crypt.
NODE_STORAGE_SECURITY_AUTHENTICATED_ENCRYPTED = 3;
}