Files
3x-ui/internal/web/service/xray_setting_dns_routing.go
T
Sanaei 892c06c8bc Bug-label issue sweep: 16 fixes (#6083)
* fix(xray): block private-range egress in default freedom finalRules (#6037)

With domainStrategy AsIs the router never resolves domains, so a domain
with a private A record (e.g. 127-0-0-1.nip.io) sails past the
geoip:private routing block and freedom's allow-all finalRules let it
reach loopback services such as the xray gRPC API and metrics listener.

Prepend a block rule for geoip:private to the default template and add
the FreedomFinalRulesPrivateEgressBlock seeder so existing installs
still carrying the stock allow-only (or legacy private-only-allow)
finalRules are upgraded in place; customized rules are left untouched.

* fix(sub): version-gate unencrypted-outbound drops in outbound subscriptions (#6033)

Commit d38c912d taught CheckXrayConfig to keep unencrypted vless/trojan
outbounds when the running core predates the v26.7.11 rejection, but
filterOutboundsRejectedByCore still consulted the embedded validator
unconditionally, so outbound subscriptions kept silently dropping those
outbounds even on downgraded cores.

Apply the same shouldSkipLegacyUnencryptedOutboundRejection gate when
filtering fetched subscription outbounds.

* fix(xray): resolve geodata assets before building outbound configs (#5928)

Saving routing or template settings validates each outbound through the
embedded config loader, and a freedom outbound whose finalRules
reference geoip:private opens geoip.dat during that build. Unlike
ApplyRoutingConfig, ValidateOutboundConfig and AddOutbound never pointed
the in-process loader at the bin folder, so xray-core resolved the file
relative to the panel executable and saving failed with
'stat /usr/local/x-ui/geoip.dat: no such file or directory'.

Call ensureXrayAssetLocation before both build paths.

* fix(api): use a real i18n key in the client get handler (#5911)

The client fetch endpoint localized its error prefix with the bare key
'get', which exists in no translation file, so every lookup of a deleted
client's email logged 'message "get" not found in language ...' noise
alongside the expected record-not-found warning. Reuse the same
pages.inbounds.toasts.obtain key the sibling list handler uses.

* fix(sub): carry host record Host header and path into Clash/JSON output (#5944)

The raw-link path overrides the host/path share params from a Host
record via applyEndpointHostPath, but the Clash and JSON renderers read
the transport settings object, which applyHostStreamOverrides never
touched — so a Host record's WebSocket Host header (and path) silently
vanished from Clash/Mihomo and JSON subscriptions whenever the inbound's
own ws settings left them empty.

Inject hostHeader/path into the ws/httpupgrade/xhttp settings of the
per-host stream, mirroring the raw-link override.

* fix(metrics): accept Unicode outbound tags in the observatory (#5972)

The observatory validator whitelisted ASCII word characters, so any
outbound whose tag carries a flag emoji or other non-ASCII text was
silently dropped from the metrics snapshot, delay history, and health
notifications. The history store is an in-process map, so the strict
charset bought nothing.

Validate tags as non-empty, bounded, control-character-free UTF-8
instead, keeping spaces and emoji while still rejecting garbage input on
the query path.

* fix(database): default sqlite to WAL to stop background-job lock storms (#6057, #6068)

With journal_mode=DELETE every write serializes the whole database and
blocks readers, so under normal multi-job load (traffic sampling, node
sync, mtproto reconcile) transactions regularly outwaited the 10s busy
timeout and jobs failed with 'database is locked'.

Move to WAL by default: readers no longer block writers and vice versa,
which removes the observed contention while writer-writer access still
serializes safely. The single-file-at-rest property is preserved where
it matters — Checkpoint() now issues wal_checkpoint(TRUNCATE), so panel
and Telegram backups read a complete main file, and sqlite folds the WAL
back into the db on clean shutdown. XUI_DB_JOURNAL_MODE=DELETE restores
the previous behavior for setups that copy the live file directly.

* fix(database): strip finalmask.tcp from REALITY inbounds on upgrade (#6038)

validateFinalMaskRealityCombo blocks saving finalmask.tcp together with
REALITY because that combination crashes Xray-core 26.7.11 on the first
connection (XTLS/Xray-core#6453), but it only runs on add/update. An
inbound saved before the validator existed sailed through the upgrade
untouched and took the core down at boot.

Add the InboundRealityFinalmaskTcpStrip seeder: one-time scan that
removes finalmask.tcp from REALITY inbounds (other finalmask transports
survive), so upgraded panels start cleanly.

* fix(xray): stop deleting hand-written direct routing rules on save (#6056)

The DNS allow-rule sync recognized 'its' rules purely by shape
(type=field, ip, port, outboundTag=direct, nothing else), so any manual
rule of that shape — e.g. routing a LAN CIDR to a NAS port over direct —
was silently stripped on every settings save.

Mark managed rules with ruleTag=xui-dns-allow (round-tripped untouched
by both xray-core and the Routing tab editor) and only strip rules that
carry the tag. Untagged legacy managed rules are adopted when their
exact ip-set/port matches a currently configured private DNS endpoint;
anything else is left alone. A stale pre-tag managed rule whose DNS
server was removed now lingers until deleted manually — the safe side of
the trade against eating user rules.

* fix(clients): resolve email lookups through client_inbounds after a move (#6059)

GetClientInboundByEmail trusted the client_traffics.inbound_id pointer
whenever that inbound still existed, but a client moved between inbounds
leaves the row pointing at its old (still existing) inbound. The lookup
then searched the wrong inbound's clients and failed with 'Client Not
Found In Inbound For Email', which broke the Telegram bot's link and QR
generation for moved clients.

When the pointed-at inbound no longer carries the email, re-resolve
through the authoritative client_inbounds link to the inbound that
actually hosts the client.

* fix(nodes): replicate inbound fallbacks to nodes (#5963)

Fallbacks live in the inbound_fallbacks table and were only merged into
settings by the master's local config builder; the runtime inbound
pushed to nodes rebuilt settings without them, and the reconcile job
additionally fingerprinted the raw DB row, so fallback edits neither
reached nodes nor triggered a re-push.

Inject settings.fallbacks in buildRuntimeInboundForAPI (mirroring the
local builder, gated on inboundCanHostFallbacks) and make ReconcileNode
push and fingerprint that same runtime-built payload, aligning the
interactive and reconcile paths.

* fix(database): survive PostgreSQL outages without a runaway restart loop (#6023)

A PostgreSQL that was down or still starting made InitDB fail instantly;
the process exited with a generic startup error and systemd restarted it
every 5s forever, flooding the journal.

Retry the initial postgres connection with backoff (~70s total) and log
the real driver error on every attempt, and cap the systemd units with
StartLimitIntervalSec/StartLimitBurst so a persistently unreachable
database stops the unit instead of looping indefinitely.

* fix(xray): force a full restart when REALITY stream settings change (#6010)

A changed inbound is normally hot-swapped over gRPC as RemoveInbound +
AddInbound, but xray-core does not reliably rebuild a REALITY listener's
authenticator on a runtime re-add — key or shortId edits appeared
applied yet clients kept authenticating against the old parameters until
someone restarted the core manually, on nodes in particular.

Treat any non-client change to an inbound that uses (or starts using)
REALITY as not hot-appliable so the panel restarts the core instead.
Client-only edits on REALITY inbounds keep flowing through the per-user
AlterInbound path and still avoid restarts.

* feat(sub): allow insecure TLS for outbound subscription fetches (#6067)

An outbound subscription served over HTTPS with a self-signed or
private-CA certificate could never be fetched: the fetch client had no
TLS options, so refreshes died with 'x509: certificate signed by unknown
authority' and there was nothing the admin could toggle.

Add a per-subscription 'Allow insecure' switch (persisted as
allow_insecure, default off) that sets InsecureSkipVerify on the fetch
transport — including when the fetch is routed through the panel egress
proxy. The SSRF-guarded dialer and redirect re-validation stay in force
either way.

* fix(reality): send PROXY protocol header in the target scanner when xver is set (#6082)

The REALITY target scanner always probed with a plain TLS handshake, so
a target fronted by an Nginx listener that requires the PROXY protocol
(matching the inbound's xver>=1) reset the connection and the panel
reported a false 'TLS handshake failed'.

Thread the inbound's xver into the scan request and, when it is >=1,
lead with the matching PROXY protocol header (v1 for xver 1, binary v2
for xver 2) built from the dialed connection's own address pair. Batch
candidate scans against public sites are unaffected (xver 0).

* fix(frontend): default sockopt fields when editing a stored inbound (#5956)

Opening an existing inbound ran rawInboundToFormValues over the raw DB
row, and only xhttpSettings was re-parsed through its Zod schema to fill
defaults. A sockopt object saved before the TProxy control existed has
no tproxy key, so the Select rendered blank; picking Off didn't help
because the wire normalizer drops tproxy=off, recreating the missing
key on the next edit.

Re-parse streamSettings.sockopt through SockoptStreamSettingsSchema on
load, mirroring the xhttpSettings handling, so absent keys (tproxy,
tcpcongestion, …) get their schema defaults every time the form opens.
2026-07-23 15:34:42 +02:00

417 lines
13 KiB
Go

package service
import (
"encoding/json"
"net"
"reflect"
"slices"
"sort"
"strconv"
"strings"
"github.com/mhsanaei/3x-ui/v3/internal/util/netsafe"
)
// dnsAllowRuleTag marks the routing rules this file manages. Both
// xray-core and the Routing tab's rule editor round-trip ruleTag
// untouched, so it is a stable provenance marker: only rules carrying it
// (or legacy managed rules recognized by exact content, see
// managedDnsAllowRule) are ever stripped and rebuilt.
const dnsAllowRuleTag = "xui-dns-allow"
// managedDnsAllowRule reports whether a routing rule was created by this
// file. Current managed rules carry dnsAllowRuleTag. Rules written before
// the tag existed are adopted only when their shape matches AND their
// exact ip-set/port content equals a currently configured private DNS
// endpoint group — a hand-written rule that merely resembles the managed
// shape (e.g. a CIDR allow for a NAS on port 5000, #6056) never matches
// and is left untouched.
func managedDnsAllowRule(rule map[string]any, groups []dnsAllowPortGroup) bool {
if tag, _ := rule["ruleTag"].(string); tag == dnsAllowRuleTag {
return true
}
if !dnsAllowRuleShape(rule) {
return false
}
port, _ := rule["port"].(string)
ips := append([]string(nil), readRuleIPs(rule["ip"])...)
sort.Strings(ips)
for _, g := range groups {
if strconv.Itoa(g.port) == port && slices.Equal(ips, g.ips) {
return true
}
}
return false
}
// dnsAllowRuleShape identifies the legacy managed-rule shape: a plain
// "type=field, ip=[...], port=..., outboundTag=direct" rule with no other
// matchers. An "enabled" key is tolerated as long as it's true — the
// Routing tab's rule editor (RuleFormModal.tsx submit()) and its enabled
// switch (RoutingTab.tsx toggleRule()) always write that key back, even
// when nothing else changed, so requiring its absence would disown the
// rule the first time an admin so much as opens it in the UI. A rule
// toggled off (enabled=false) is treated as no longer ours: the admin
// explicitly turned it off, and re-enabling it on the next save would
// silently override that choice.
func dnsAllowRuleShape(rule map[string]any) bool {
if t, _ := rule["type"].(string); t != "field" {
return false
}
if out, _ := rule["outboundTag"].(string); out != "direct" {
return false
}
if _, ok := rule["ip"]; !ok {
return false
}
if _, ok := rule["port"]; !ok {
return false
}
for key := range rule {
switch key {
case "type", "outboundTag", "ip", "port":
continue
case "enabled":
if enabled, ok := rule[key].(bool); !ok || !enabled {
return false
}
continue
default:
return false
}
}
return true
}
// findPrivateBlockRule returns the index of a routing rule that blocks
// geoip:private (the panel's default anti-SSRF rule), or -1 if none is
// present. Matched by shape (outboundTag=blocked, ip contains
// "geoip:private") rather than position, since admins can reorder rules.
func findPrivateBlockRule(rules []map[string]any) int {
for i, rule := range rules {
if out, _ := rule["outboundTag"].(string); out != "blocked" {
continue
}
for _, ip := range readRuleIPs(rule["ip"]) {
if strings.EqualFold(ip, "geoip:private") {
return i
}
}
}
return -1
}
func readRuleIPs(raw any) []string {
switch v := raw.(type) {
case []string:
return v
case []any:
out := make([]string, 0, len(v))
for _, item := range v {
if s, ok := item.(string); ok {
out = append(out, s)
}
}
return out
case string:
if v == "" {
return nil
}
return []string{v}
default:
return nil
}
}
// dnsServerEndpoint is a literal (ip, port) pair extracted from a
// dns.servers entry.
type dnsServerEndpoint struct {
ip string
port int
}
// privateDnsServerEndpoint extracts a literal, private/internal (ip, port)
// endpoint from a dns.servers entry, or ok=false if the entry is a domain
// name, a special Xray keyword (localhost, fakedns, ...), or resolves to a
// public IP.
//
// A dns.servers entry is either a bare string or an object with an
// "address" field (see frontend/src/schemas/dns.ts DnsServerEntrySchema);
// the object form may also carry an explicit "port" (default 53 there,
// per DnsServerObjectInnerSchema), which takes precedence over any port
// embedded in the address itself.
func privateDnsServerEndpoint(entry any) (dnsServerEndpoint, bool) {
var address string
explicitPort := 0
switch v := entry.(type) {
case string:
address = v
case map[string]any:
address, _ = v["address"].(string)
if p, ok := v["port"].(float64); ok && p > 0 {
explicitPort = int(p)
}
default:
return dnsServerEndpoint{}, false
}
host, port := splitAddressHostPort(address)
if host == "" {
return dnsServerEndpoint{}, false
}
if explicitPort > 0 {
port = explicitPort
}
ip := net.ParseIP(host)
if ip == nil {
// Domain name, or a special keyword like "localhost"/"fakedns" —
// neither is something we can safely allow-list by IP here.
return dnsServerEndpoint{}, false
}
if !netsafe.IsBlockedIP(ip) {
return dnsServerEndpoint{}, false
}
return dnsServerEndpoint{ip: ip.String(), port: port}, true
}
// splitAddressHostPort extracts the bare host and port (defaulting to 53)
// from an Xray-core DNS server address string. Those may carry a URI
// scheme (tcp://, tcp+local://, https://, https+local://, quic://,
// quic+local://) and, for DoH, a path and/or a bracketed IPv6 host — all
// of that is stripped down to host[:port] before parsing.
func splitAddressHostPort(address string) (host string, port int) {
address = strings.TrimSpace(address)
if address == "" {
return "", 0
}
if idx := strings.Index(address, "://"); idx != -1 {
address = address[idx+3:]
}
// Drop a DoH path, e.g. "1.1.1.1/dns-query".
if idx := strings.Index(address, "/"); idx != -1 {
address = address[:idx]
}
port = 53
host = address
if strings.HasPrefix(host, "[") {
// Bracketed IPv6, with or without a port: "[::1]" / "[::1]:53".
end := strings.Index(host, "]")
if end == -1 {
return host, port
}
rest := host[end+1:]
host = host[1:end]
if p, ok := strings.CutPrefix(rest, ":"); ok {
if n, err := strconv.Atoi(p); err == nil {
port = n
}
}
return host, port
}
if h, p, err := net.SplitHostPort(host); err == nil {
host = h
if n, err := strconv.Atoi(p); err == nil {
port = n
}
}
return host, port
}
// dnsAllowPortGroup is the set of private literal IPs that share a single
// port among the configured dns.servers, e.g. two internal resolvers both
// queried on :53.
type dnsAllowPortGroup struct {
port int
ips []string
}
// collectPrivateDnsAllowGroups returns the private dns.servers endpoints
// grouped by port, sorted by port ascending (ips within a group sorted and
// de-duplicated) for deterministic output.
func collectPrivateDnsAllowGroups(dnsRaw json.RawMessage) []dnsAllowPortGroup {
if len(dnsRaw) == 0 {
return nil
}
var dns struct {
Servers []any `json:"servers"`
}
if err := json.Unmarshal(dnsRaw, &dns); err != nil {
return nil
}
byPort := make(map[int]map[string]bool)
for _, entry := range dns.Servers {
ep, ok := privateDnsServerEndpoint(entry)
if !ok {
continue
}
if byPort[ep.port] == nil {
byPort[ep.port] = make(map[string]bool)
}
byPort[ep.port][ep.ip] = true
}
ports := make([]int, 0, len(byPort))
for p := range byPort {
ports = append(ports, p)
}
sort.Ints(ports)
groups := make([]dnsAllowPortGroup, 0, len(ports))
for _, p := range ports {
ips := make([]string, 0, len(byPort[p]))
for ip := range byPort[p] {
ips = append(ips, ip)
}
sort.Strings(ips)
groups = append(groups, dnsAllowPortGroup{port: p, ips: ips})
}
return groups
}
// EnsureDnsServerRouting keeps a set of managed "direct" allow-rules — one
// per distinct port among any private/internal dns.servers addresses —
// in sync, positioned immediately before the panel's default
// geoip:private block rule.
//
// Why this matters: Xray's own DNS client traffic is dispatched through
// the same routing table as proxied client traffic. If dns.servers points
// at a private IP (e.g. a self-hosted AdGuard Home / Pi-hole reachable on
// the same Docker network as Xray — a common self-hosted setup) and the
// panel's default private-IP block rule is active, Xray's own DNS lookups
// get silently dropped by that rule. Xray then falls back to dialing
// destinations by raw hostname once its internal DNS attempt times out
// (~4s), so proxied connections still work, just with a multi-second stall
// added to every new domain, with no error surfaced to the client or
// admin.
//
// Each managed rule is scoped to its port (not just the IP), so the
// exception only reopens the DNS traffic that actually needs it rather
// than every port on the private host. On every save, all previously
// managed rules are stripped out and a fresh set is rebuilt from the
// current dns.servers config and reinserted right before the block rule
// (recomputing its index after the strip) — this corrects both content
// drift (dns.servers changed) and position drift (an admin dragged a
// managed rule below the block rule in the Routing tab, which would
// otherwise silently reintroduce the stall with nothing to notice or fix
// it). The rebuilt result is only written back if it actually differs
// from the input, so well-formed configs aren't churned on every save.
// Manually-authored rules are never touched — see managedDnsAllowRule.
func EnsureDnsServerRouting(raw string) (string, error) {
var cfg map[string]json.RawMessage
if err := json.Unmarshal([]byte(raw), &cfg); err != nil {
return raw, err
}
groups := collectPrivateDnsAllowGroups(cfg["dns"])
var routing map[string]json.RawMessage
if r, ok := cfg["routing"]; ok && len(r) > 0 {
if err := json.Unmarshal(r, &routing); err != nil {
return raw, err
}
}
if routing == nil {
return raw, nil
}
var original []map[string]any
if r, ok := routing["rules"]; ok && len(r) > 0 {
if err := json.Unmarshal(r, &original); err != nil {
return raw, err
}
}
rebuilt := rebuildDnsAllowRules(original, groups)
rulesJSON, err := json.Marshal(rebuilt)
if err != nil {
return raw, err
}
// Compare against the original rules JSON, not the parsed Go values:
// json.Unmarshal into []map[string]any turns "ip" arrays into []any,
// while the rules this function builds use []string — those hold
// identical content but are different types under reflect.DeepEqual,
// which would otherwise report a no-op input as changed and churn the
// JSON on every save for no reason.
origRulesJSON := routing["rules"]
if len(origRulesJSON) == 0 {
origRulesJSON = json.RawMessage("[]")
}
if jsonEqual(origRulesJSON, rulesJSON) {
return raw, nil
}
routing["rules"] = rulesJSON
routingJSON, err := json.Marshal(routing)
if err != nil {
return raw, err
}
cfg["routing"] = routingJSON
out, err := json.Marshal(cfg)
if err != nil {
return raw, err
}
return string(out), nil
}
// rebuildDnsAllowRules strips any existing managed rules out of rules,
// then — if a geoip:private block rule is present and groups is non-empty
// — reinserts a freshly built managed rule per group immediately before
// it. This uniformly handles content updates, position drift, and removal
// (an empty groups list just leaves the managed rules stripped).
func rebuildDnsAllowRules(rules []map[string]any, groups []dnsAllowPortGroup) []map[string]any {
clean := make([]map[string]any, 0, len(rules))
for _, rule := range rules {
if !managedDnsAllowRule(rule, groups) {
clean = append(clean, rule)
}
}
blockIdx := findPrivateBlockRule(clean)
if blockIdx < 0 || len(groups) == 0 {
return clean
}
managed := make([]map[string]any, 0, len(groups))
for _, g := range groups {
managed = append(managed, map[string]any{
"type": "field",
"ruleTag": dnsAllowRuleTag,
"ip": g.ips,
"port": strconv.Itoa(g.port),
"outboundTag": "direct",
})
}
// Capacity hint uses len(clean) alone (not len(clean)+len(managed)):
// summing two independent lengths for a make() size risks overflow on
// pathological input per static analysis, and clean's length already
// covers most of the eventual size on its own.
out := make([]map[string]any, 0, len(clean))
out = append(out, clean[:blockIdx]...)
out = append(out, managed...)
out = append(out, clean[blockIdx:]...)
return out
}
// jsonEqual reports whether a and b decode to structurally identical
// values. Used instead of comparing raw bytes (key order, whitespace) or
// reflect.DeepEqual on already-parsed Go values (which is type-sensitive
// to []any vs []string and would misreport identical content as changed).
func jsonEqual(a, b json.RawMessage) bool {
var av, bv any
if err := json.Unmarshal(a, &av); err != nil {
return false
}
if err := json.Unmarshal(b, &bv); err != nil {
return false
}
return reflect.DeepEqual(av, bv)
}