Files
seaweedfs/weed/admin/dash/admin_data.go
T
Chris Lu 8236c8d136 vector: integrate engine health, admin management, and local deployment (#12517)
* vector: expose request metrics and registration-aware health endpoints

* vector: report immutable read cache and storage fetch metrics

* docs: describe vector engine health, metrics, and maintenance

* vector: count canceled requests when their handler is dropped

* vector: account for uncached large immutable storage reads

* vector: reliably release the stalled credential reader in tests

* vector: observe large reads as their streamed body is consumed

* vector: bound heartbeats on established master registration streams

* test: exercise hardened vector maintenance on Go and Rust volumes

* vector: register engines in their configured filer group

* vector: make retried index creation preserve compatible datasets

* vector: allow registration to advertise a separate address

* test: release stalled credential readers without blocking FIFO cleanup

* test: allow hardened vector coverage in a named filer group

* vector: discover engines in the same filer group as their gateway

* vector: fall back when the advertised address is empty

* admin: expose vector readiness, maintenance, and management tools (#12520)

* s3vectors: share vector discovery with trusted management clients

* admin: show vector engine readiness and recent maintenance activity

* admin: suggest S3 Vectors IAM actions from the supported operations

* admin: manage vector indexes, policies, tags, encryption, and query data

* admin: validate SSE-S3 support before creating encrypted vector buckets

* admin: discover vector engines in their registration group

* admin: probe vector engines concurrently with independent time budgets

* admin: verify encryption capability on the remote S3 gateway

* admin: defer discovery until after the HTTP server can start

* admin: distinguish unknown vector health services from failed engines

* admin: count workers that can execute vector index maintenance

* admin: allow read-only users to query vectors and inspect tags

* vector: scope gateway and admin discovery to the configured filer group

* admin: refresh policy state after mutations and serialize policy controls

* vector: drain active RPCs before closing replaced discovery connections

* vector: respect request cancellation while waiting for discovery

* vector: prefer another registered engine after a transport failure

* admin: check all masters before reporting no registered vector engines

* iam: honor attached vector policies without broadening legacy grants

* vector: allow short-lived discovery clients to close their connections

* vector: refresh cached master addresses for each discovery lookup

* iam: enforce vector bucket tag conditions in attached policies

* iam: authorize vector bucket creation with the requested tags

* test: exercise vector IAM grants and tag denies on Go and Rust volumes

* admin: reject authenticated vector mutations without an admin role

* test: run vector IAM assertions before the shared engine is stopped

* vector: add native mini startup, shell tools, and a runnable Compose example (#12522)

* s3vectors: share vector discovery with trusted management clients

* admin: show vector engine readiness and recent maintenance activity

* admin: suggest S3 Vectors IAM actions from the supported operations

* admin: manage vector indexes, policies, tags, encryption, and query data

* admin: validate SSE-S3 support before creating encrypted vector buckets

* admin: discover vector engines in their registration group

* admin: probe vector engines concurrently with independent time budgets

* admin: verify encryption capability on the remote S3 gateway

* admin: defer discovery until after the HTTP server can start

* admin: distinguish unknown vector health services from failed engines

* admin: count workers that can execute vector index maintenance

* admin: allow read-only users to query vectors and inspect tags

* vector: scope gateway and admin discovery to the configured filer group

* admin: refresh policy state after mutations and serialize policy controls

* vector: drain active RPCs before closing replaced discovery connections

* vector: respect request cancellation while waiting for discovery

* vector: prefer another registered engine after a transport failure

* admin: check all masters before reporting no registered vector engines

* iam: honor attached vector policies without broadening legacy grants

* vector: allow short-lived discovery clients to close their connections

* vector: refresh cached master addresses for each discovery lookup

* iam: enforce vector bucket tag conditions in attached policies

* iam: authorize vector bucket creation with the requested tags

* test: exercise vector IAM grants and tag denies on Go and Rust volumes

* admin: reject authenticated vector mutations without an admin role

* test: run vector IAM assertions before the shared engine is stopped

* mini: optionally start and supervise the vector engine

* mini: bootstrap vector buckets with existing Lance catalog helpers

* shell: manage vector buckets, indexes, and vector data

* docs: provide a runnable vector and maintenance worker Compose stack

* shell: scope vector discovery to the session and close its connections

* mini: honor the bind address for vector and metrics listeners

* mini: reserve vector ports before assigning gRPC and compatibility listeners

* mini: wait for vector storage setup and master registration

* mini: create requested vector buckets without requiring the Lance HTTP listener

* shell: refresh master addresses during vector discovery

* mini: resolve sibling vector executables with platform extensions

* vector: preserve uncertain creates and fix mini listener allocation (#12525)

* vector: retain catalog records when index creation has an unknown outcome

* mini: keep chosen listener ports reserved throughout allocation

* mini: resolve hostname binds before starting vector listeners

* vector: fix integration test imports and recover interrupted index creation (#12533)

* shell: register vector operations from the command package

* s3vectors: resume interrupted index initialization safely

* mini: exclude disabled WebDAV from listener allocation

* s3vectors: retain uncertain creates across refused retries

* s3vectors: claim pending indexes before retrying initialization

* s3vectors: compare resolved distance metrics during recovery

* s3vectors: preserve metadata keys when retrying initialization

* s3vectors: verify ownership before initializing new catalog records
2026-11-10 04:21:18 -08:00

477 lines
16 KiB
Go

package dash
import (
"context"
"net"
"net/http"
"sort"
"time"
"github.com/seaweedfs/seaweedfs/weed/cluster"
"github.com/seaweedfs/seaweedfs/weed/glog"
"github.com/seaweedfs/seaweedfs/weed/iam"
"github.com/seaweedfs/seaweedfs/weed/pb"
"github.com/seaweedfs/seaweedfs/weed/pb/master_pb"
)
// Access key status constants
const (
AccessKeyStatusActive = iam.AccessKeyStatusActive
AccessKeyStatusInactive = iam.AccessKeyStatusInactive
)
type AdminData struct {
Username string `json:"username"`
TotalVolumes int `json:"total_volumes"`
TotalChunks int64 `json:"total_chunks"`
TotalSize int64 `json:"total_size"`
DiskUsage int64 `json:"disk_usage"`
DiskCapacity int64 `json:"disk_capacity"`
VolumeSizeLimitMB uint64 `json:"volume_size_limit_mb"`
MasterNodes []MasterNode `json:"master_nodes"`
VolumeServers []VolumeServer `json:"volume_servers"`
FilerNodes []FilerNode `json:"filer_nodes"`
MessageBrokers []MessageBrokerNode `json:"message_brokers"`
VectorNodes []VectorNode `json:"vector_nodes"`
S3Nodes []S3Node `json:"s3_nodes"`
DataCenters []DataCenter `json:"datacenters"`
LastUpdated time.Time `json:"last_updated"`
// EC shard totals for dashboard
TotalEcVolumes int `json:"total_ec_volumes"` // Total number of EC volumes across all servers
TotalEcShards int `json:"total_ec_shards"` // Total number of EC shards across all servers
// TotalMountClients is the number of connected FUSE/VFS mount clients across filers.
TotalMountClients int `json:"total_mount_clients"`
// Trends holds at-a-glance sparklines built from the admin's own recent
// cluster snapshots (no Prometheus required).
Trends DashboardTrends `json:"trends"`
// TierStats breaks volumes and EC shards down by storage tier: local
// disk types plus one entry per remote storage holding tiered volumes.
TierStats []TierStats `json:"tier_stats"`
// VersionSummary reports whether all cluster nodes run the same binary
// version, so the dashboard can warn about partially upgraded clusters.
VersionSummary ClusterVersionSummary `json:"version_summary"`
}
// Object Store Users management structures
type ObjectStoreUser struct {
Username string `json:"username"`
Email string `json:"email"`
AccessKey string `json:"access_key"`
SecretKey string `json:"secret_key"`
Permissions []string `json:"permissions"`
PolicyNames []string `json:"policy_names"`
IsStatic bool `json:"is_static"` // loaded from static config file, not editable
}
type ObjectStoreUsersData struct {
Username string `json:"username"`
Users []ObjectStoreUser `json:"users"`
TotalUsers int `json:"total_users"`
HasAnonymousUser bool `json:"has_anonymous_user"`
LastUpdated time.Time `json:"last_updated"`
}
// User management request structures
type CreateUserRequest struct {
Username string `json:"username" binding:"required"`
Email string `json:"email"`
Actions []string `json:"actions"`
GenerateKey bool `json:"generate_key"`
PolicyNames []string `json:"policy_names"`
}
type UpdateUserRequest struct {
Email string `json:"email"`
Actions []string `json:"actions"`
PolicyNames []string `json:"policy_names"`
}
type UpdateUserPoliciesRequest struct {
Actions []string `json:"actions" binding:"required"`
}
type AccessKeyInfo struct {
AccessKey string `json:"access_key"`
SecretKey string `json:"secret_key"`
Status string `json:"status"`
CreatedAt time.Time `json:"created_at"`
}
type CreateAccessKeyRequest struct {
AccessKey string `json:"access_key"`
SecretKey string `json:"secret_key"`
}
type UpdateAccessKeyStatusRequest struct {
Status string `json:"status" binding:"required"`
}
type UserDetails struct {
Username string `json:"username"`
Email string `json:"email"`
Actions []string `json:"actions"`
PolicyNames []string `json:"policy_names"`
AccessKeys []AccessKeyInfo `json:"access_keys"`
Groups []string `json:"groups"`
}
// RoleReadOnly is the session role assigned to read-only (view-only) admin
// accounts. It matches the value stored by HandleLogin when the read-only
// credentials are used.
const RoleReadOnly = "readonly"
// IsReadOnlyRole reports whether the given admin session role grants only
// view-only access. Any other role (admin, or the empty role used when auth
// is disabled) is treated as non-read-only.
func IsReadOnlyRole(role string) bool {
return role == RoleReadOnly
}
// RedactSecretKey clears the plaintext S3 secret key from an object-store
// user record. The access key (a public identifier) is retained so the
// identity can still be listed; only the reusable secret is removed.
func (u *ObjectStoreUser) RedactSecretKey() {
u.SecretKey = ""
}
// RedactSecretKeys clears the plaintext S3 secret keys from a user's access
// key records. Access key identifiers are retained so the set of keys remains
// visible; only the reusable secrets are removed.
func (d *UserDetails) RedactSecretKeys() {
for i := range d.AccessKeys {
d.AccessKeys[i].SecretKey = ""
}
}
type FilerNode struct {
Address string `json:"address"`
DataCenter string `json:"datacenter"`
Rack string `json:"rack"`
Version string `json:"version"`
LastUpdated time.Time `json:"last_updated"`
}
type MessageBrokerNode struct {
Address string `json:"address"`
DataCenter string `json:"datacenter"`
Rack string `json:"rack"`
Version string `json:"version"`
LastUpdated time.Time `json:"last_updated"`
}
type S3Node struct {
Address string `json:"address"`
DataCenter string `json:"datacenter"`
Version string `json:"version"`
LastUpdated time.Time `json:"last_updated"`
}
// GetAdminData retrieves admin data as a struct (for reuse by both JSON and HTML handlers)
func (s *AdminServer) GetAdminData(username string) (AdminData, error) {
if username == "" {
username = "admin"
}
// Get cluster topology
topology, err := s.GetClusterTopology()
if err != nil {
glog.Errorf("Failed to get cluster topology: %v", err)
return AdminData{}, err
}
// Get volume servers data with EC shard information
volumeServersData, err := s.GetClusterVolumeServers()
if err != nil {
glog.Errorf("Failed to get cluster volume servers: %v", err)
return AdminData{}, err
}
// Get master nodes status
masterNodes := s.getMasterNodesStatus()
// Get filer nodes status
filerNodes := s.getFilerNodesStatus()
// Get message broker nodes status
messageBrokers := s.getMessageBrokerNodesStatus()
// Get S3 nodes status
s3Nodes := s.getS3NodesStatus()
// Get volume size limit from master configuration
var volumeSizeLimitMB uint64 = 30000 // Default to 30GB
err = s.WithMasterClient(func(client master_pb.SeaweedClient) error {
resp, err := client.GetMasterConfiguration(context.Background(), &master_pb.GetMasterConfigurationRequest{})
if err != nil {
return err
}
volumeSizeLimitMB = uint64(resp.VolumeSizeLimitMB)
return nil
})
if err != nil {
glog.Warningf("Failed to get volume size limit from master: %v", err)
// Keep default value on error
}
// Calculate EC shard totals
var totalEcVolumes, totalEcShards int
ecVolumeSet := make(map[uint32]bool) // To avoid counting the same EC volume multiple times
for _, vs := range volumeServersData.VolumeServers {
totalEcShards += vs.EcShards
// Count unique EC volumes across all servers
for _, ecInfo := range vs.EcShardDetails {
ecVolumeSet[ecInfo.VolumeID] = true
}
}
totalEcVolumes = len(ecVolumeSet)
// Count connected FUSE/VFS mount clients (best-effort; don't fail the dashboard)
totalMountClients := 0
if mountData, mountErr := s.GetMountClients(); mountErr == nil {
totalMountClients = mountData.TotalMountClients
}
// Prepare admin data
adminData := AdminData{
Username: username,
TotalVolumes: topology.TotalVolumes,
TotalChunks: topology.TotalChunks,
TotalSize: topology.TotalSize,
DiskUsage: topology.DiskUsage,
DiskCapacity: topology.DiskCapacity,
VolumeSizeLimitMB: volumeSizeLimitMB,
MasterNodes: masterNodes,
VolumeServers: volumeServersData.VolumeServers,
FilerNodes: filerNodes,
MessageBrokers: messageBrokers,
S3Nodes: s3Nodes,
VectorNodes: s.getVectorNodesStatus(context.Background()).Engines,
DataCenters: topology.DataCenters,
LastUpdated: topology.UpdatedAt,
TotalEcVolumes: totalEcVolumes,
TotalEcShards: totalEcShards,
TotalMountClients: totalMountClients,
Trends: s.GetDashboardTrends(),
TierStats: topology.TierStats,
VersionSummary: buildClusterVersionSummary(volumeServersData.VolumeServers, filerNodes, s3Nodes, messageBrokers),
}
return adminData, nil
}
// ShowAdmin displays the main admin page (now uses GetAdminData)
func (s *AdminServer) ShowAdmin(w http.ResponseWriter, r *http.Request) {
username := UsernameFromContext(r.Context())
adminData, err := s.GetAdminData(username)
if err != nil {
writeJSONError(w, http.StatusInternalServerError, "Failed to get admin data: "+err.Error())
return
}
// Return JSON for API calls
writeJSON(w, http.StatusOK, adminData)
}
// ShowOverview displays cluster overview
func (s *AdminServer) ShowOverview(w http.ResponseWriter, r *http.Request) {
topology, err := s.GetClusterTopology()
if err != nil {
writeJSONError(w, http.StatusInternalServerError, err.Error())
return
}
writeJSON(w, http.StatusOK, topology)
}
// getMasterNodesStatus returns the full set of master nodes in the cluster.
// It prefers the authoritative raft membership (RaftListClusterServers) and
// falls back to the currently-connected master if the raft call fails, so the
// dashboard never shows an empty list.
func (s *AdminServer) getMasterNodesStatus() []MasterNode {
masterMap := make(map[string]MasterNode)
raftReturnedEmpty := false
err := s.WithMasterClient(func(client master_pb.SeaweedClient) error {
ctx, cancel := context.WithTimeout(context.Background(), 3*time.Second)
defer cancel()
resp, err := client.RaftListClusterServers(ctx, &master_pb.RaftListClusterServersRequest{})
if err != nil {
return err
}
raftReturnedEmpty = len(resp.ClusterServers) == 0
for _, server := range resp.ClusterServers {
// Skip malformed raft addresses instead of letting an
// unconvertible value into masterMap.
if _, _, splitErr := net.SplitHostPort(server.Address); splitErr != nil {
glog.Warningf("skip master with invalid raft address %q: %v", server.Address, splitErr)
continue
}
httpAddress := pb.GrpcAddressToServerAddress(server.Address)
if httpAddress == "" {
glog.Warningf("skip master with invalid raft address %q", server.Address)
continue
}
masterMap[httpAddress] = MasterNode{
Address: httpAddress,
IsLeader: server.IsLeader,
}
}
return nil
})
if err != nil {
currentMaster := s.masterClient.GetMaster(context.Background())
glog.Errorf("Failed to list raft cluster masters from %s: %v", currentMaster, err)
}
if len(masterMap) == 0 {
currentMaster := s.masterClient.GetMaster(context.Background())
if currentMaster != "" {
addr := pb.ServerAddress(currentMaster).ToHttpAddress()
// A successful empty raft response means raft is not initialized
// (standalone/non-raft cluster); the only master IS the leader.
// A failed RPC or a nonempty response whose entries were all
// rejected must not claim leadership.
masterMap[addr] = MasterNode{
Address: addr,
IsLeader: raftReturnedEmpty,
}
}
}
masterNodes := make([]MasterNode, 0, len(masterMap))
for _, m := range masterMap {
masterNodes = append(masterNodes, m)
}
sort.Slice(masterNodes, func(i, j int) bool {
return masterNodes[i].Address < masterNodes[j].Address
})
return masterNodes
}
// getFilerNodesStatus checks status of all filer nodes using master's ListClusterNodes
func (s *AdminServer) getFilerNodesStatus() []FilerNode {
var filerNodes []FilerNode
// Get filer nodes from master using ListClusterNodes
err := s.WithMasterClient(func(client master_pb.SeaweedClient) error {
resp, err := client.ListClusterNodes(context.Background(), s.listClusterNodesRequest(cluster.FilerType))
if err != nil {
return err
}
// Process each filer node
for _, node := range resp.ClusterNodes {
filerNodes = append(filerNodes, FilerNode{
Address: pb.ServerAddress(node.Address).ToHttpAddress(),
DataCenter: node.DataCenter,
Rack: node.Rack,
Version: node.Version,
LastUpdated: time.Now(),
})
}
return nil
})
if err != nil {
currentMaster := s.masterClient.GetMaster(context.Background())
glog.Errorf("Failed to get filer nodes from master %s: %v", currentMaster, err)
// Return empty list if we can't get filer info from master
return []FilerNode{}
}
// Sort filer nodes by address for consistent ordering on page refresh
sort.Slice(filerNodes, func(i, j int) bool {
return filerNodes[i].Address < filerNodes[j].Address
})
return filerNodes
}
// getMessageBrokerNodesStatus checks status of all message broker nodes using master's ListClusterNodes
func (s *AdminServer) getMessageBrokerNodesStatus() []MessageBrokerNode {
var messageBrokers []MessageBrokerNode
// Get message broker nodes from master using ListClusterNodes
err := s.WithMasterClient(func(client master_pb.SeaweedClient) error {
resp, err := client.ListClusterNodes(context.Background(), s.listClusterNodesRequest(cluster.BrokerType))
if err != nil {
return err
}
// Process each message broker node
for _, node := range resp.ClusterNodes {
messageBrokers = append(messageBrokers, MessageBrokerNode{
Address: node.Address,
DataCenter: node.DataCenter,
Rack: node.Rack,
Version: node.Version,
LastUpdated: time.Now(),
})
}
return nil
})
if err != nil {
currentMaster := s.masterClient.GetMaster(context.Background())
glog.Errorf("Failed to get message broker nodes from master %s: %v", currentMaster, err)
// Return empty list if we can't get broker info from master
return []MessageBrokerNode{}
}
// Sort message broker nodes by address for consistent ordering on page refresh
sort.Slice(messageBrokers, func(i, j int) bool {
return messageBrokers[i].Address < messageBrokers[j].Address
})
return messageBrokers
}
// getS3NodesStatus checks status of all S3 nodes using master's ListClusterNodes
func (s *AdminServer) getS3NodesStatus() []S3Node {
var s3Nodes []S3Node
// Get S3 nodes from master using ListClusterNodes
err := s.WithMasterClient(func(client master_pb.SeaweedClient) error {
resp, err := client.ListClusterNodes(context.Background(), s.listClusterNodesRequest(cluster.S3Type))
if err != nil {
return err
}
// Process each S3 node
for _, node := range resp.ClusterNodes {
s3Nodes = append(s3Nodes, S3Node{
Address: pb.ServerAddress(node.Address).ToHttpAddress(),
DataCenter: node.DataCenter,
Version: node.Version,
LastUpdated: time.Now(),
})
}
return nil
})
if err != nil {
currentMaster := s.masterClient.GetMaster(context.Background())
glog.Errorf("Failed to get S3 nodes from master %s: %v", currentMaster, err)
// Return empty list if we can't get S3 info from master
return []S3Node{}
}
// Sort S3 nodes by address for consistent ordering on page refresh
sort.Slice(s3Nodes, func(i, j int) bool {
return s3Nodes[i].Address < s3Nodes[j].Address
})
return s3Nodes
}