feat: rework engine and CLI for dnstraverse parity
Port the traversal engine to the Ruby dnstraverse model so behaviour and output match dns.squish.net: - dns: single RD=0 query path (RD=1 only for upstream root discovery), per-run packet cache, EDNS0 512-fallback with warnings, UDP->TCP on truncation; fix --retries 0 and --root-server IP-literal handling; drop all hardcoded 127.0.0.1:53 resolvers - traverse: hierarchical per-branch InfoCache, 7-step response classification with the full 10-status vocabulary, bailiwick partitioning, strictly-deeper lame-referral rule, refid grammar with .0 resolve subtrees and childset digits, per-IP branching at 1/n weight, cache-based glue resolution with noglue/loop dead ends, CNAME restarts from the deepest cached zone, fast-mode memoization, probability aggregation with Ruby-identical stats keys (sums to 1.0) - output: byte-for-byte reference text format pinned by a golden test, reference CLI defaults, working --quiet/--show-X=false, TTY-aware colour, deduplicated deterministic JSON - web: adapt API/SPA to the new engine, SSE events carry refid/status, fix subscribe/snapshot duplicate-event race and a statusCls TDZ bug, align SPA type list with the backend - delete the old engine and dead code (net -4,350 lines) Verified against live runs of the reference Ruby engine across five domains (answers, NXDOMAIN, null MX, CNAME restart, glueless resolve) with no divergences beyond the documented typo fixes. Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
This commit is contained in:
Gary Hansen
co-authored by
Claude Fable 5
Claude Fable 5
parent
af15c9c2d4
commit
d71c7fbef2
@@ -0,0 +1,187 @@
|
||||
package traverse
|
||||
|
||||
import (
|
||||
"fmt"
|
||||
"sort"
|
||||
|
||||
miekgdns "github.com/miekg/dns"
|
||||
)
|
||||
|
||||
// ServerResponse wraps a DecodedQuery for one (server, IP) query, mirroring
|
||||
// response.rb: it owns a child InfoCache seeded with the in-bailiwick records,
|
||||
// upgrades referral → referral_lame, and computes the start servers for
|
||||
// referral/restart children. The noglue/loop variants (response_noglue.rb,
|
||||
// response_loop.rb) are synthetic — no query was sent, DQ is nil.
|
||||
type ServerResponse struct {
|
||||
DQ *DecodedQuery
|
||||
Status Status
|
||||
|
||||
Qname string
|
||||
Qclass uint16
|
||||
Qtype uint16
|
||||
IP string
|
||||
Server string
|
||||
Bailiwick string
|
||||
// ParentIP is the address of the referring server; it is part of the
|
||||
// stats key for referral_lame so lame referrals from different parents
|
||||
// stay separate.
|
||||
ParentIP string
|
||||
|
||||
Cache *InfoCache
|
||||
Starters []StartServer
|
||||
StartersBailiwick string
|
||||
}
|
||||
|
||||
// NewServerResponse evaluates a decoded query in the context of parentCache.
|
||||
// server is the NS hostname that was queried (dq.IP is its address).
|
||||
func NewServerResponse(dq *DecodedQuery, server, parentIP string, parentCache *InfoCache) (*ServerResponse, error) {
|
||||
r := &ServerResponse{
|
||||
DQ: dq,
|
||||
Status: dq.Status,
|
||||
Qname: dq.Qname,
|
||||
Qclass: dq.Qclass,
|
||||
Qtype: dq.Qtype,
|
||||
IP: dq.IP,
|
||||
Server: canonicalName(server),
|
||||
Bailiwick: dq.Bailiwick,
|
||||
ParentIP: parentIP,
|
||||
Cache: NewInfoCache(parentCache),
|
||||
}
|
||||
if err := r.evaluate(); err != nil {
|
||||
return nil, err
|
||||
}
|
||||
return r, nil
|
||||
}
|
||||
|
||||
// NewNoGlueResponse records a dead end: ip referred us to server inside
|
||||
// bailiwick without glue and there is no way to resolve it.
|
||||
func NewNoGlueResponse(qname string, qclass, qtype uint16, ip, server, bailiwick string) *ServerResponse {
|
||||
return &ServerResponse{
|
||||
Status: StatusNoGlue,
|
||||
Qname: canonicalName(qname),
|
||||
Qclass: qclass,
|
||||
Qtype: qtype,
|
||||
IP: ip,
|
||||
Server: canonicalName(server),
|
||||
Bailiwick: canonicalName(bailiwick),
|
||||
}
|
||||
}
|
||||
|
||||
// NewLoopResponse records a dead end: resolving server from ip would repeat
|
||||
// an ancestor referral.
|
||||
func NewLoopResponse(qname string, qclass, qtype uint16, ip, server, bailiwick string) *ServerResponse {
|
||||
return &ServerResponse{
|
||||
Status: StatusLoop,
|
||||
Qname: canonicalName(qname),
|
||||
Qclass: qclass,
|
||||
Qtype: qtype,
|
||||
IP: ip,
|
||||
Server: canonicalName(server),
|
||||
Bailiwick: canonicalName(bailiwick),
|
||||
}
|
||||
}
|
||||
|
||||
// evaluate mirrors response.rb#evaluate: cache the in-bailiwick records, then
|
||||
// for referral/restart work out the start servers from THIS branch's cache;
|
||||
// a referral whose new zone is not strictly deeper than the bailiwick is lame.
|
||||
func (r *ServerResponse) evaluate() error {
|
||||
if r.Status != StatusException {
|
||||
r.Cache.Add(r.DQ.CacheableGood)
|
||||
}
|
||||
switch r.DQ.Status {
|
||||
case StatusRestart:
|
||||
starters, bw, err := r.Cache.GetStartServers(r.DQ.Endname)
|
||||
if err != nil {
|
||||
return err
|
||||
}
|
||||
r.Starters, r.StartersBailiwick = starters, bw
|
||||
case StatusReferral:
|
||||
starters, bw, err := r.Cache.GetStartServers(r.DQ.Endname)
|
||||
if err != nil {
|
||||
return err
|
||||
}
|
||||
r.Starters, r.StartersBailiwick = starters, bw
|
||||
if isLameReferral(r.DQ.Bailiwick, bw) {
|
||||
r.Status = StatusReferralLame
|
||||
}
|
||||
starterNames := make([]string, len(starters))
|
||||
for i, s := range starters {
|
||||
starterNames[i] = s.Name
|
||||
}
|
||||
if !equalSorted(starterNames, r.DQ.AuthorityNames) {
|
||||
r.DQ.WarningsAdd("Referred authority names do not match query cache expectations")
|
||||
}
|
||||
}
|
||||
return nil
|
||||
}
|
||||
|
||||
func equalSorted(a, b []string) bool {
|
||||
if len(a) != len(b) {
|
||||
return false
|
||||
}
|
||||
as := append([]string(nil), a...)
|
||||
bs := append([]string(nil), b...)
|
||||
sort.Strings(as)
|
||||
sort.Strings(bs)
|
||||
for i := range as {
|
||||
if as[i] != bs[i] {
|
||||
return false
|
||||
}
|
||||
}
|
||||
return true
|
||||
}
|
||||
|
||||
// StatsKey is the leaf aggregation key (response.rb update_stats_key and the
|
||||
// noglue/loop variants): identical keys merge by summing probability.
|
||||
func (r *ServerResponse) StatsKey() string {
|
||||
qclass := ClassToString(r.Qclass)
|
||||
qtype := TypeToString(r.Qtype)
|
||||
switch r.Status {
|
||||
case StatusNoGlue, StatusLoop:
|
||||
return fmt.Sprintf("key:%s:%s:%s:%s:%s:%s:%s",
|
||||
r.Status, r.IP, r.Qname, qclass, qtype, r.Server, r.Bailiwick)
|
||||
default:
|
||||
key := fmt.Sprintf("key:%s:%s:%s:%s:%s:%s",
|
||||
r.Status, r.IP, r.Server, r.Qname, qclass, qtype)
|
||||
if r.Status == StatusException && r.DQ != nil {
|
||||
key += ":" + r.DQ.ExceptionMessage
|
||||
} else if r.Status == StatusReferralLame {
|
||||
key += ":" + r.ParentIP
|
||||
}
|
||||
return key
|
||||
}
|
||||
}
|
||||
|
||||
// String renders a short description for progress display (the Ruby
|
||||
// response to_s variants: "No glue for X" / "Loop encountered resolving X").
|
||||
func (r *ServerResponse) String() string {
|
||||
switch r.Status {
|
||||
case StatusNoGlue:
|
||||
return fmt.Sprintf("No glue for %s", r.Server)
|
||||
case StatusLoop:
|
||||
return fmt.Sprintf("Loop encountered resolving %s", r.Server)
|
||||
case StatusException:
|
||||
if r.DQ != nil {
|
||||
return r.DQ.ExceptionMessage
|
||||
}
|
||||
case StatusError:
|
||||
if r.DQ != nil {
|
||||
return r.DQ.ErrorMessage
|
||||
}
|
||||
}
|
||||
return string(r.Status)
|
||||
}
|
||||
|
||||
func ClassToString(qclass uint16) string {
|
||||
if s, ok := miekgdns.ClassToString[qclass]; ok {
|
||||
return s
|
||||
}
|
||||
return fmt.Sprintf("CLASS%d", qclass)
|
||||
}
|
||||
|
||||
func TypeToString(qtype uint16) string {
|
||||
if s, ok := miekgdns.TypeToString[qtype]; ok {
|
||||
return s
|
||||
}
|
||||
return fmt.Sprintf("TYPE%d", qtype)
|
||||
}
|
||||
Reference in New Issue
Block a user