Phase 2b: refactor to file-backed storage; UPDATE writes zones/*.zone

Major architectural pivot per the user's "RFC 2136 mechanism for the
existing zonefiles, not a new in-memory thing" framing. The plugin no
longer maintains its own in-memory state OR serves any queries -- both
of those are now the auto plugin's job, reading the same zone files.

The plugin's sole responsibility is now: receive TSIG-authed UPDATE
messages, edit the matching zones/<zone>.zone file, bump the SOA
serial in CalVer (YYYYMMDDNN) form, and optionally auto-commit to git.

What changed:
- DELETED: store.go (in-memory recordStore), store_test.go (12 tests),
  plugin_test.go (10 ServeDNS query tests), old update_test.go.
- NEW: zonefile.go -- file-backed authority for one zone. loadRRs via
  miekg/dns zone parser; mutation helpers (lookupIn/nameExistsIn/
  removeRRsetFrom/removeRRFrom/removeNameFrom/addRRTo) on []dns.RR
  slices; bumpSerial with CalVer semantics + NN exhaustion handling;
  writeAtomic via temp-file rename; commit shells to `git add && git
  commit` with configurable author.
- NEW: zonefile_test.go -- 17 tests covering load/lookup/mutate/bump/
  write paths.
- REWRITTEN: plugin.go -- ServeDNS is now thin: UPDATE → TSIG → handler;
  everything else → Next. No synthetic SOA/NS, no query serving.
- REWRITTEN: update.go -- handleUpdate now opens the zoneFile, loads,
  applies (with prereq checks against the loaded RRs), bumps serial,
  writes, commits. Detects no-op updates to avoid spurious file writes.
- REWRITTEN: setup.go -- new directives: `zones-dir` (required),
  `auto-commit` (default true), `git-author <name> <email>`. Dropped
  `nameserver` and `persist`. Validates each declared zone has a file
  on disk via os.Stat before CoreDNS finishes starting.
- REWRITTEN: setup_test.go -- 17 cases for the new grammar.
- REWRITTEN: update_test.go -- 11 cases using real temp zone files
  via t.TempDir().

Total: 30 tests passing, 0 failures.

Next: Phase 2c (custom CoreDNS image, deploy, smoke test with nsupdate).
This commit is contained in:
Ryan Malloy 2026-05-21 11:26:50 -06:00
parent 1d2d919728
commit 0f28127284
10 changed files with 1223 additions and 1109 deletions

196
plugin.go
View file

@ -1,61 +1,67 @@
// Package rfc2136 implements a CoreDNS plugin that accepts dynamic DNS
// updates per RFC 2136 (UPDATE opcode), authenticated via TSIG. The
// primary use case is self-hosted ACME DNS-01 cert automation: an ACME
// client (e.g. Caddy via caddy-dns/rfc2136) injects _acme-challenge TXT
// records into a delegated sub-zone that this plugin serves.
// Package rfc2136 is a CoreDNS plugin that accepts dynamic DNS updates
// per RFC 2136 (UPDATE opcode), authenticated via TSIG, and applies
// them to on-disk zone files. This is the right shape for stacks where
// the operator wants to keep zones in flat files (perhaps under git,
// with HE pulling AXFR), but also wants programmatic updates from
// clients like Caddy's caddy-dns/rfc2136 module.
//
// Phase 1.3 status: store + query dispatch. ServeDNS now answers
// authoritatively for the configured zone(s) from the in-memory store
// (plus synthetic SOA/NS at apex). UPDATE handling still rejected —
// that lands in Phase 1.4. See plan at
// The plugin does NOT serve any queries — that's the job of the
// `auto`/`file` plugin running alongside it. This plugin's only
// responsibility is the UPDATE opcode path: verify TSIG, dissect the
// UPDATE, write the zone file, bump the SOA serial, optionally
// auto-commit to git. CoreDNS's auto plugin notices the mtime change
// and re-serves the zone within its reload interval.
//
// ~/.claude/plans/dood-does-coredns-offer-enumerated-piglet.md
// See the plan at
// ~/.claude/plans/dood-does-coredns-offer-enumerated-piglet.md
// for the architectural rationale.
package rfc2136
import (
"context"
"strings"
"github.com/coredns/coredns/plugin"
"github.com/miekg/dns"
)
// DefaultTTL is the TTL applied to dynamically-added records when the
// Corefile doesn't specify one. 60s matches the short-lived nature of
// ACME challenge TXT records and keeps stale answers from lingering in
// DefaultTTL is applied to dynamically-added records whose UPDATE
// messages carry TTL=0. 60s matches the short-lived nature of ACME
// challenge records and keeps stale answers from lingering in
// resolver caches.
const DefaultTTL uint32 = 60
// RFC2136 is the plugin handler. One instance per Corefile server block.
type RFC2136 struct {
// Next is the downstream plugin in the chain.
// Next is the downstream plugin in the chain — queries always
// pass through; only UPDATE opcode is intercepted.
Next plugin.Handler
// Zones is the set of canonical (dot-terminated, lowercase) zone
// names this instance is authoritative for. Queries outside these
// zones pass through to Next.
// names this instance accepts UPDATEs for. UPDATEs for any other
// zone are rejected with NOTAUTH.
Zones []string
// TSIGKeys is keyed by canonical key name (lowercased, trailing
// dot). Empty means TSIG is disabled — UPDATEs without TSIG are
// rejected unconditionally in Phase 1.4.
// dot). Empty means TSIG is disabled — UPDATEs are refused
// unconditionally as a safety default.
TSIGKeys map[string]tsigKey
// TTL is applied to dynamically-injected records that don't carry
// an explicit TTL in the UPDATE message.
TTL uint32
// PersistPath, when non-empty, names a file the plugin writes a
// JSON snapshot of its in-memory store to on a periodic schedule.
// Empty means in-memory only (acceptable for ACME challenges).
PersistPath string
// ZonesDir is the directory where <zone>.zone files live (matching
// the mount path inside the CoreDNS container). The plugin reads
// and writes files at <ZonesDir>/<zone>.zone.
ZonesDir string
// Nameserver is the host returned in synthetic NS records and as
// the SOA's MNAME. Defaults (set in setup) to the first zone apex.
Nameserver string
// AutoCommit governs whether the plugin auto-commits zone-file
// changes to git after every successful UPDATE.
AutoCommit bool
// store holds the dynamic records. Always non-nil after setup.
store *recordStore
// zones holds per-zone file handlers, keyed by canonical zone name.
// Populated in setup; mutexes live inside each zoneFile.
zones map[string]*zoneFile
}
// Name implements plugin.Handler.
@ -65,19 +71,11 @@ func (p *RFC2136) Name() string { return "rfc2136" }
//
// Dispatch:
//
// UPDATE opcode → rejected with REFUSED (Phase 1.4 implements properly).
// Query opcode:
// - Not in our zones → pass to Next.
// - Apex SOA → synthetic SOA.
// - Apex NS → synthetic NS.
// - Match in store → return RRset.
// - Name exists, wrong type → NODATA (NOERROR + SOA in authority).
// - Name doesn't exist → NXDOMAIN (NameError + SOA in authority).
// UPDATE opcode → verify TSIG, then apply via the UPDATE handler.
// Anything else → pass through to Next (the auto plugin handles
// queries against the zone files we maintain).
func (p *RFC2136) ServeDNS(ctx context.Context, w dns.ResponseWriter, r *dns.Msg) (int, error) {
if r.Opcode == dns.OpcodeUpdate {
// TSIG verification was performed by the underlying dns.Server
// (because setup.go populated dnsserver.Config.TsigSecret). We
// just need to check the result here.
if err := p.checkTSIG(w, r); err != nil {
log.Warningf("UPDATE rejected: %v", err)
resp := new(dns.Msg)
@ -87,123 +85,5 @@ func (p *RFC2136) ServeDNS(ctx context.Context, w dns.ResponseWriter, r *dns.Msg
}
return p.handleUpdate(w, r)
}
if len(r.Question) == 0 {
return plugin.NextOrFailure(p.Name(), p.Next, ctx, w, r)
}
q := r.Question[0]
zone := p.findZone(q.Name)
if zone == "" {
return plugin.NextOrFailure(p.Name(), p.Next, ctx, w, r)
}
// We're authoritative for this name. Build a reply.
msg := new(dns.Msg)
msg.SetReply(r)
msg.Authoritative = true
qname := strings.ToLower(dns.Fqdn(q.Name))
isApex := qname == zone
// Apex SOA / NS are synthetic.
if isApex {
switch q.Qtype {
case dns.TypeSOA:
msg.Answer = []dns.RR{p.syntheticSOA(zone)}
_ = w.WriteMsg(msg)
return dns.RcodeSuccess, nil
case dns.TypeNS:
msg.Answer = p.syntheticNS(zone)
_ = w.WriteMsg(msg)
return dns.RcodeSuccess, nil
}
}
// Look up the asked type in the store.
if rrs := p.store.Lookup(qname, q.Qtype); rrs != nil {
msg.Answer = rrs
_ = w.WriteMsg(msg)
return dns.RcodeSuccess, nil
}
// Special case: ANY at the apex isn't in the store but we have
// synthetic SOA + NS. Return them rather than NODATA.
if isApex && q.Qtype == dns.TypeANY {
msg.Answer = append(msg.Answer, p.syntheticSOA(zone))
msg.Answer = append(msg.Answer, p.syntheticNS(zone)...)
_ = w.WriteMsg(msg)
return dns.RcodeSuccess, nil
}
// Distinguish NODATA from NXDOMAIN.
if p.store.NameExists(qname) || isApex {
// NODATA: name exists, but not with this type.
msg.Ns = []dns.RR{p.syntheticSOA(zone)}
_ = w.WriteMsg(msg)
return dns.RcodeSuccess, nil
}
// NXDOMAIN: name doesn't exist anywhere in this zone.
msg.Rcode = dns.RcodeNameError
msg.Ns = []dns.RR{p.syntheticSOA(zone)}
_ = w.WriteMsg(msg)
return dns.RcodeNameError, nil
}
// findZone returns the longest matching zone for qname, or "" if qname
// is outside all configured zones. The returned zone is in canonical
// form (lowercase, trailing dot).
func (p *RFC2136) findZone(qname string) string {
qname = strings.ToLower(dns.Fqdn(qname))
// Longest-suffix wins so nested zones work correctly.
var best string
for _, z := range p.Zones {
if qname == z || strings.HasSuffix(qname, "."+z) {
if len(z) > len(best) {
best = z
}
}
}
return best
}
// syntheticSOA returns the SOA RR for a zone. Serial is derived from
// the store's monotonic generation counter — every UPDATE bumps it,
// so downstream observers can detect "something changed" without
// having to AXFR.
func (p *RFC2136) syntheticSOA(zone string) *dns.SOA {
return &dns.SOA{
Hdr: dns.RR_Header{
Name: zone,
Rrtype: dns.TypeSOA,
Class: dns.ClassINET,
Ttl: p.TTL,
},
Ns: p.Nameserver,
Mbox: "admin." + zone,
Serial: uint32(p.store.generation()),
Refresh: 3600, // 1 hour
Retry: 600, // 10 min
Expire: 604800, // 1 week
Minttl: 60, // negative-cache TTL
}
}
// syntheticNS returns the NS RRset for a zone. Currently a single NS
// pointing at p.Nameserver (the host that runs this plugin). For
// resiliency, future versions could accept multiple `nameserver`
// directives.
func (p *RFC2136) syntheticNS(zone string) []dns.RR {
return []dns.RR{
&dns.NS{
Hdr: dns.RR_Header{
Name: zone,
Rrtype: dns.TypeNS,
Class: dns.ClassINET,
Ttl: p.TTL,
},
Ns: p.Nameserver,
},
}
return plugin.NextOrFailure(p.Name(), p.Next, ctx, w, r)
}