Files
developer 504512f247 Add a Zones console: create and manage local zones, records and delegations
The console could not see local zones at all. `blocklist.local_zones` compiled
a store that nothing read back: there was no API for a zone, its records, its
delegations or the file a record came from, and no way to change one without
editing YAML and reloading. The Domains page is the blocklist's inline rules,
not local data, so nothing in the console covered the half of the server that
answers authoritatively from disk. There is now a Zones page and the endpoints
behind it: `GET /api/zones`, `POST`/`PUT`/`DELETE /api/zones/records`,
`POST`/`DELETE /api/zones/delegations` and `DELETE /api/zones?zone=`.

**The console owns one directory and nothing else.** `blocklist.managed_zones`
holds one `<apex>.zone` per zone, with that zone's SOA, apex NS set and the
records added there, and it is compiled by the same `localzone.Load` as the
operator's files — so a record added from the console is served on the next
query, with no reload and no restart. Files in `blocklist.local_zones` are
listed and shown read-only and are never rewritten; the key defaults to empty,
so a deployment that never sets it gets a zone browser and no zone writes.
Which file a record came from is now part of the store (`zone.files`, keyed by
the index the store already uses), so the snapshot can say "console" or "your
file" per record; the query path does not read it.

**The writer proves its own output.** A record is validated by the project's own
master-file parser, so `www` means `www.<zone>` and a bare CNAME target is
completed exactly as it would be in a zone file; the file is rendered, and the
rendering is re-parsed and compared with the records it meant to write before
the file is renamed into place. A writer bug that dropped a record would
otherwise present as a zone that silently stopped answering. A zone that cannot
answer is refused rather than created: no SOA means no NXDOMAIN or NODATA
authority, and no apex NS means nothing is delegated and no server is named. The
SOA's primary follows the apex NS set — a zone whose SOA names a server outside
its own NS set hands out a referral no resolver can follow — and the serial
moves on every write.

**A delegation takes the names under it, so a record below a cut is refused.**
The store answers those names as delegated without reading the zone file, which
would make the record data that looks alive and never answers. `DELETE
/api/zones?zone=` exists for the same reason from the other side: a zone the
console created is a real zone, so it says NXDOMAIN for what it does not hold,
and deleting a record is not what gives a name back to upstream. It is
refused for a zone that also has records outside the managed directory, because
removing that file would remove none of the zone.

**Two bugs the work turned up.** `engine.applyConfig` published local-record
metrics only when `blocklist.local_zones` was non-empty, so a zone the console
created was served but never counted; the gate is now the store's own record
count. And the record-value renderer fell back to `RR.String()` for every type
it did not know, which is the whole line rather than the RDATA — it wrote a
name and a TTL into the value column for DHCID, LOC, SSHFP and SVCB, and would
have fed that back as a record value on the next edit.

Every write recompiles the store and flushes the cache. Which names changed is
not knowable from the API, and a cached upstream answer for a name the zone now
answers is simply wrong, so the flush is blunt on purpose.

The wire shapes are type aliases in `internal/api` rather than copies, so there
is one set of json tags, and `TestZoneEndpointKeySets` pins the key set at every
level with a record and a delegation actually present — an empty list only
proves the envelope. `TestZoneWritesNeedAManagedDirectory` pins the read-only
default, and `TestZoneConsoleRoundTrip` walks the whole thing: a record added
through the API is answered authoritatively on the next query, a cut stops the
parent answering below it, and deleting the zone hands the name back.

Verified against a running instance built from this tree: records and
wildcards answer with `aa`, relative names are qualified, a record below a cut
is refused with a message that says why, and the file on disk is a zone file
an operator can read. The console page is typechecked and built; the browser was
not available in this round, so it has not been looked at.
2026-09-28 07:45:16 +00:00

301 lines
8.9 KiB
Go

package localzone
import (
"fmt"
"path/filepath"
"sort"
"strconv"
"strings"
"github.com/miekg/dns"
)
// Record is one authoritative record as the API and the console see it.
//
// The compiled store keeps plain dns.RR slices, because that is what the
// query path reads; this presentation form is built on demand for the admin
// API only, so a zone browser can show what the operator actually wrote.
type Record struct {
// Name is the fully-qualified owner, trailing dot included. A wildcard
// owner is reported as written ("*.example.com.").
Name string `json:"name"`
// Type is the RR type, uppercase ("A", "MX", "NS").
Type string `json:"type"`
// TTL is the record's own TTL in seconds.
TTL uint32 `json:"ttl"`
// Value is the presentation-format RDATA with every domain name
// absolute, so a record read out of the store can be pasted into a new
// zone file unchanged.
Value string `json:"value"`
// Wildcard is true when the owner was written "*.name".
Wildcard bool `json:"wildcard"`
// Delegation is true for an NS record below the apex: the cut that hands
// the names under it to another server.
Delegation bool `json:"delegation"`
// Managed is true when the record came from the console-managed
// directory, the only data the API may rewrite.
Managed bool `json:"managed"`
// Source is the file the record was parsed from.
Source string `json:"source"`
}
// Key identifies the RRset a record belongs to: owner name plus type. A
// delegation is the NS RRset of one name, so it is the same key.
func (r Record) Key() string { return r.Name + "|" + r.Type }
// Zone is a compiled zone as the admin API reports it.
type Zone struct {
// Origin is the apex, lowercased and fully qualified.
Origin string `json:"origin"`
// Serial is the SOA serial, 0 when the zone has no SOA.
Serial uint32 `json:"serial"`
// Soa is the zone's start of authority in presentation form, with every
// name absolute, or "" when the zone has no SOA. It is not a record the
// API writes: the writer owns it, and a new managed file is seeded from
// this so a zone keeps the server list it already had.
Soa string `json:"soa"`
// DefaultTTL is the zone's smallest TTL, which is what a record written
// without one inherits.
DefaultTTL uint32 `json:"default_ttl"`
// Records are sorted by owner, then type, then value.
Records []Record `json:"records"`
// Delegations is the subset of Records that is a cut below the apex,
// with the NS targets collected per name.
Delegations []Delegation `json:"delegations"`
// RecordCount is len(Records).
RecordCount int `json:"record_count"`
// Managed is true when every source file of the zone is inside the
// console-managed directory, i.e. the whole zone is API-writable.
Managed bool `json:"managed"`
// Files are the sources this zone was compiled from.
Files []string `json:"files"`
}
// Delegation is a cut in the zone: a name below the apex whose NS records
// hand it, and everything under it, to other servers.
type Delegation struct {
// Name is the delegated name, fully qualified.
Name string `json:"name"`
// Targets are the NS targets, lowercased and fully qualified.
Targets []string `json:"targets"`
// Managed is true when the delegation came from the managed directory.
Managed bool `json:"managed"`
}
// Snapshot renders one zone for the API, or nil when the store holds no such
// zone. Records from every source file are included and each one carries the
// file it came from, because the console has to be able to tell an
// operator's read-only master file apart from data it may rewrite.
func (s *Store) Snapshot(origin, managedDir string) *Zone {
if s == nil {
return nil
}
z, ok := s.zones[strings.ToLower(dns.Fqdn(origin))]
if !ok {
return nil
}
view := &Zone{
Origin: z.origin,
DefaultTTL: z.minTTL,
Files: append([]string(nil), z.sources...),
Managed: true,
}
if z.soa != nil {
view.Serial = z.soa.Serial
view.Soa = fmt.Sprintf("%s %s %d %d %d %d %d", z.soa.Ns, z.soa.Mbox,
z.soa.Serial, z.soa.Refresh, z.soa.Retry, z.soa.Expire, z.soa.Minttl)
}
for k, rrs := range z.exact {
name, rrtype, _ := strings.Cut(k, "|")
for _, rr := range rrs {
view.Records = append(view.Records, z.record(rr, name, rrtype, k, managedDir))
}
}
for name, byType := range z.wild {
for rrtype, rrs := range byType {
rrtypeStr := dns.TypeToString[rrtype]
k := wildKey(name, rrtypeStr)
for _, rr := range rrs {
view.Records = append(view.Records, z.record(rr, "*."+name, rrtypeStr, k, managedDir))
}
}
}
for _, rec := range view.Records {
if !rec.Managed {
view.Managed = false
break
}
}
sortRecords(view.Records)
view.RecordCount = len(view.Records)
view.Delegations = delegations(view.Records)
return view
}
// SnapshotAll renders every compiled zone, sorted by apex.
func (s *Store) SnapshotAll(managedDir string) []*Zone {
if s == nil {
return nil
}
out := make([]*Zone, 0, len(s.zones))
for _, origin := range s.Zones() {
if z := s.Snapshot(origin, managedDir); z != nil {
out = append(out, z)
}
}
return out
}
// record builds the presentation form of one compiled record. key is the
// index it is stored under, which is where its source file is recorded.
func (z *zone) record(rr dns.RR, name, rrtype, key, managedDir string) Record {
rec := Record{
Name: name,
Type: rrtype,
TTL: rr.Header().Ttl,
Value: rdata(rr),
Wildcard: strings.HasPrefix(name, "*."),
Source: z.source(key),
}
// The apex NS set is the zone's own server list, not a cut; only an NS
// record below the apex hands names to another server.
rec.Delegation = rec.Type == "NS" && !rec.Wildcard && name != z.origin
rec.Managed = inDir(rec.Source, managedDir)
return rec
}
// delegations collects the cuts out of a rendered record list.
func delegations(recs []Record) []Delegation {
byName := make(map[string]*Delegation)
var order []string
for _, rec := range recs {
if !rec.Delegation {
continue
}
d, ok := byName[rec.Name]
if !ok {
d = &Delegation{Name: rec.Name, Managed: rec.Managed}
byName[rec.Name] = d
order = append(order, rec.Name)
}
d.Targets = append(d.Targets, rec.Value)
}
sort.Strings(order)
out := make([]Delegation, 0, len(order))
for _, name := range order {
d := byName[name]
sort.Strings(d.Targets)
d.Targets = dedupe(d.Targets)
out = append(out, *d)
}
return out
}
// rdata renders the RDATA of a record with domain names absolute. The generic
// fallback is the RR's own String(), which is right for every type not listed
// here.
func rdata(rr dns.RR) string {
switch v := rr.(type) {
case *dns.A:
return v.A.String()
case *dns.AAAA:
return v.AAAA.String()
case *dns.CNAME:
return v.Target
case *dns.NS:
return v.Ns
case *dns.PTR:
return v.Ptr
case *dns.MX:
return utoa(uint64(v.Preference)) + " " + v.Mx
case *dns.SRV:
return utoa(uint64(v.Priority)) + " " + utoa(uint64(v.Weight)) + " " +
utoa(uint64(v.Port)) + " " + v.Target
case *dns.TXT:
return quoteStrings(v.Txt)
case *dns.CAA:
return utoa(uint64(v.Flag)) + " " + v.Tag + " " + quote(v.Value)
default:
// RR.String() is the whole line, owner and TTL included. The RDATA
// is everything after the type, which is the fourth tab-separated
// field.
fields := strings.Split(rr.String(), "\t")
if len(fields) < 5 {
return strings.TrimSpace(rr.String())
}
return strings.Join(fields[4:], " ")
}
}
func utoa(v uint64) string { return strconv.FormatUint(v, 10) }
func quoteStrings(in []string) string {
out := make([]string, 0, len(in))
for _, s := range in {
out = append(out, quote(s))
}
return strings.Join(out, " ")
}
func quote(s string) string {
var sb strings.Builder
sb.Grow(len(s) + 2)
sb.WriteByte('"')
for i := 0; i < len(s); i++ {
if c := s[i]; c == '"' || c == '\\' {
sb.WriteByte('\\')
}
sb.WriteByte(s[i])
}
sb.WriteByte('"')
return sb.String()
}
// sortRecords orders a rendered zone by owner, then type, then value, so the
// console's table is stable and the writer's output is deterministic.
func sortRecords(recs []Record) {
sort.Slice(recs, func(i, j int) bool {
if recs[i].Name != recs[j].Name {
return recs[i].Name < recs[j].Name
}
if recs[i].Type != recs[j].Type {
return recs[i].Type < recs[j].Type
}
return recs[i].Value < recs[j].Value
})
}
func dedupe(in []string) []string {
if len(in) == 0 {
return in
}
out := in[:1]
for _, s := range in[1:] {
if s != out[len(out)-1] {
out = append(out, s)
}
}
return out
}
// inDir reports whether a source file lives under a directory, which is how a
// record is recognised as console-managed.
func inDir(file, dir string) bool {
if file == "" || dir == "" {
return false
}
abs, err := filepath.Abs(file)
if err != nil {
abs = file
}
base, err := filepath.Abs(dir)
if err != nil {
base = dir
}
rel, err := filepath.Rel(base, abs)
if err != nil {
return false
}
return rel != ".." && !strings.HasPrefix(rel, ".."+string(filepath.Separator))
}