Files
seaweedfs/weed/admin/dash/admin_data.go
T
Chris Lu 37bf1cd91d volume: validate copy/tail source addresses before dialing (#11390)
* pb: stop exiting the process on malformed server addresses

ServerToGrpcAddress and GrpcAddressToServerAddress called glog.Fatalf
when hostAndPort could not parse the port, which os.Exit(255)ed the whole
process. A caller-supplied copy or tail source address reached this path
synchronously in the serving goroutine, so one anonymous VolumeCopy with
a non-numeric port terminated the volume server.

Log the parse error and return the input unchanged instead: the dial or
request that consumes the address then fails as an ordinary error.

* volume: validate copy and tail source addresses before dialing

VolumeCopy, VolumeEcShardsCopy and VolumeTailReceiver dial a
caller-supplied source address (SourceDataNode / SourceVolumeServer)
with no endpoint validation, so an anonymous caller could aim the volume
server at loopback, link-local (cloud metadata) or other unintended
destinations and read dial behavior back as a connectivity oracle.

Apply the same peer-target deny list FetchAndWriteNeedle uses for
replica targets: the source must be a bare host:port whose host is not
loopback, link-local or unspecified; cluster peers stay reachable on
private networks, and -volume.allowUntrustedRemoteEndpoints opts out.
The loopback-using copy tests set the flag to keep exercising the copy
path in process.

* rust volume: validate copy and tail source addresses before dialing

Mirror the Go guard on the Rust volume server: volume_copy,
volume_ec_shards_copy and volume_tail_receiver dial a caller-supplied
source address, so run it through validate_replica_target first (bare
host:port; no loopback, link-local or unspecified hosts; private peers
stay allowed). --volume.allowUntrustedRemoteEndpoints opts out; the test
fixture and the Rust test-cluster launcher set it so loopback sources in
tests keep working.

* volume: pin validated copy/tail source addresses at dial time

validateReplicaTarget resolves the source hostname once, but the gRPC
client resolved it again at connect, leaving a DNS-rebinding window for
hostname sources. The copy and tail source dials now run through the
same guardedDialerPolicy the remote-storage path uses, so every resolved
address is re-checked against the replica deny list (private peers
allowed) immediately before the TCP connect. guardedDialerPolicy also
moves to util.OutboundDialContext so the guarded path keeps the -ip.bind
source binding the default gRPC dialer had.

The Rust volume server mirrors this with connect_guarded, a tonic
connector that resolves, re-checks each address, and connects to the
first passing IP; handlers use it whenever the untrusted-endpoint
opt-out is off. A handler-level test now exercises the enabled
validation branches for all three source-taking RPCs.

* pb: return empty server address for malformed grpc addresses

GrpcAddressToServerAddress used to return the unparseable input on a
hostAndPort failure, so a malformed raft address (e.g. "host:abc")
flowed into admin dashboard master maps unchanged. Return an empty
string instead, skip empty conversions at the two raft-cluster merge
sites, and drop the now-stale comment about the fatal exit the earlier
commit removed.

* test: opt erasure-coding loopback clusters out of the remote endpoint guard

The erasure-coding suites drive VolumeEcShardsCopy / VolumeCopy between
volume servers bound to 127.0.0.1, which the copy/tail source guard now
rejects by default. Pass -volume.allowUntrustedRemoteEndpoints to the
test volume launches, matching what the volume_server framework
harnesses already do.

* admin: only claim fallback master leadership on an empty raft response

A nonempty RaftListClusterServers response whose entries were all
rejected left masterMap empty, so the fallback marked the reachable
current master as leader the same way a genuinely empty (non-raft)
response does. Track whether the successful response returned zero
servers and only promote the fallback master then.
2026-09-18 12:55:47 -07:00

460 lines
15 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"`
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"`
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"`
}
// 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"`
LastUpdated time.Time `json:"last_updated"`
}
type MessageBrokerNode struct {
Address string `json:"address"`
DataCenter string `json:"datacenter"`
Rack string `json:"rack"`
LastUpdated time.Time `json:"last_updated"`
}
type S3Node struct {
Address string `json:"address"`
DataCenter string `json:"datacenter"`
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,
VolumeSizeLimitMB: volumeSizeLimitMB,
MasterNodes: masterNodes,
VolumeServers: volumeServersData.VolumeServers,
FilerNodes: filerNodes,
MessageBrokers: messageBrokers,
S3Nodes: s3Nodes,
DataCenters: topology.DataCenters,
LastUpdated: topology.UpdatedAt,
TotalEcVolumes: totalEcVolumes,
TotalEcShards: totalEcShards,
TotalMountClients: totalMountClients,
Trends: s.GetDashboardTrends(),
TierStats: topology.TierStats,
}
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,
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,
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,
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
}