inbound-antispam-panel: refused inbound mail from mail.log; retention

Step 4/12 of docs/plans/inbound-antispam-panel.md.

The log tailer writes an inbound_spam_log row for mail port 25 refused
(internal/logtail/inbound.go): a "reject:" line of the port-25 smtpd is
Postfix's own policy; a milter-reject / milter-hold / milter-discard
line of cleanup is the filter's verdict. Only a recipient at an inbound
domain is a row, so "Relay access denied" for a stranger's domain is
not. The row is dated by its log line; the reply text is cut to 255
bytes of printable ASCII; a reply starting with 4 is a tempfail.

To tell port 25 from 465 / 587 the listener now logs as
postfix/inbound (syslog_name). cleanup logs under one name for every
listener, so its verdict counts only for a queue id the port-25 smtpd
was seen to open (4096 remembered). Lines are matched from the head,
after timestamp and host: their tail is the sender's own text.

Retention: PruneInboundSpam every six hours - older than the
inbound_spam_log_retention_days setting (seeded from
INBOUND_SPAM_LOG_RETENTION_DAYS, 90, range 7-365), then down to 10 000
rows. The setting has no field in the panel yet.

Known limits, in architecture.md: a log line names one recipient, so a
message refused at end-of-message is one row; a line re-read after an
unclean stop would be a second row.

Checked on the test server, uncommitted tree: go vet ./... ok,
go test ./... ok (22 packages), make e2e ok - the unknown recipient at
the inbound domain refused on port 25 has its row (reject by postfix),
the stranger's domain has none.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
This commit is contained in:
mix
2026-10-11 22:42:25 +03:00
co-authored by Claude Opus 5.5
parent 94e94490af
commit 27fd355e53
18 changed files with 744 additions and 25 deletions
+231
View File
@@ -0,0 +1,231 @@
package logtail
import (
"context"
"log"
"regexp"
"strings"
"time"
"github.com/mixeme/selfpost/internal/store"
)
// InboundStore is what the tailer needs of the inbound antispam journal:
// writing the rows for refused mail, knowing which domains are inbound, and
// pruning. *store.Store satisfies it.
type InboundStore interface {
InsertInboundSpam(e store.InboundSpamEntry) error
IsInboundDomain(name string) (bool, error)
PruneInboundSpam(cutoff time.Time, keep int) (int64, error)
}
// Inbound turns on the inbound half of the tailer: journal rows for mail that
// port 25 refused, and the journal's retention. Run takes nil when the inbound
// relay is off.
type Inbound struct {
Store InboundStore
Retention RetentionDays
}
// The inbound journal milter writes a row for mail that was accepted; mail
// that was refused never reaches it, and what is known about it is one line
// of mail.log:
//
// postfix/inbound/smtpd[41]: NOQUEUE: reject: RCPT from mx.example.net[203.0.113.9]: 550 5.1.1 <nobody@in.example.com>: Recipient address rejected: User unknown in relay recipient table; from=<a@example.net> to=<nobody@in.example.com> proto=ESMTP helo=<mx.example.net>
// postfix/cleanup[52]: 4A1B2C3D: milter-reject: END-OF-MESSAGE from mx.example.net[203.0.113.9]: 5.7.1 Spam message rejected; from=<a@example.net> to=<b@in.example.com> proto=ESMTP helo=<mx.example.net>
//
// The first is Postfix's own policy, written by the port-25 smtpd, which
// build/postfix-config.sh names postfix/inbound so that its lines are told
// from those of the submission ports. The second is the filter's verdict at
// end-of-message, which cleanup(8) writes under one name for every listener -
// so it counts only for a queue id the port-25 smtpd was seen to open
// (inboundClientRe).
//
// Both are matched from the head of the line, after the timestamp and host
// name: the tail of such a line is the sender's own words (HELO name,
// addresses), and a pattern that could start anywhere would let him write
// journal rows.
var (
inboundRejectRe = regexp.MustCompile(`^\S+ postfix/(inbound/smtpd|cleanup)\[\d+\]: ([0-9A-Za-z]+): (reject|milter-reject|milter-hold|milter-discard): \S+ from [^\[\s]*\[([0-9A-Fa-f:.]+)\]: (.*); from=<([^>]*)> to=<([^>]*)> proto=\S+(?: helo=<[^>]*>)?$`)
inboundClientRe = regexp.MustCompile(`^\S+ postfix/inbound/smtpd\[\d+\]: ([0-9A-Za-z]+): client=`)
)
// A row holds one line of reason; Postfix's reply text carries the far end's
// bytes and is cut and reduced to printable ASCII like any of them.
const inboundDetailMax = 255
// inboundQueueIDs is how many port-25 queue ids are remembered while waiting
// for cleanup's verdict line on them. A message is in that state for the
// seconds its DATA takes; the cap only bounds a flood.
const inboundQueueIDs = 4096
// inboundJournal is the per-run state of the inbound half.
type inboundJournal struct {
st InboundStore
// port25 holds the queue ids the port-25 smtpd opened and cleanup has not
// reported on yet; order is their arrival, for dropping the oldest.
port25 map[string]struct{}
order []string
}
func newInboundJournal(st InboundStore) *inboundJournal {
return &inboundJournal{st: st, port25: make(map[string]struct{})}
}
// handle looks at one mail.log line and writes a journal row when it is a
// refusal of mail for an inbound domain. Errors are logged; the tailer goes on.
func (j *inboundJournal) handle(line string, now time.Time) {
at, rest := lineTime(line, now)
if m := inboundClientRe.FindStringSubmatch(rest); m != nil {
j.remember(m[1])
return
}
m := inboundRejectRe.FindStringSubmatch(rest)
if m == nil {
return
}
program, queueID, kind, clientIP, reason, from, to := m[1], m[2], m[3], m[4], m[5], m[6], m[7]
if program == "cleanup" {
if _, ok := j.port25[queueID]; !ok {
return // a submission-port message, or one opened before this run
}
delete(j.port25, queueID)
}
// Only mail for an inbound domain is the relay's. "Relay access denied"
// for a stranger's domain is every scanner on the internet, and a line
// with no recipient (size limit at MAIL) belongs to no domain.
domain := ""
if i := strings.LastIndexByte(to, '@'); i >= 0 {
domain = strings.ToLower(to[i+1:])
}
if domain == "" {
return
}
ok, err := j.st.IsInboundDomain(domain)
if err != nil {
log.Printf("log-tailer: inbound journal: look up %s: %v", domain, err)
return
}
if !ok {
return
}
e := store.InboundSpamEntry{
InboundDomain: domain,
ClientIP: clientIP,
From: from,
To: to,
Decision: store.SpamDecisionReject,
Engine: store.SpamEngineRspamd,
Detail: printable(reason, inboundDetailMax),
At: at,
}
switch {
case kind == "reject":
e.Engine = store.SpamEnginePostfix
case kind == "milter-hold":
e.Decision = store.SpamDecisionQuarantine
}
// "450 4.7.1 …" from Postfix, "4.7.1 …" from a milter: either way a reply
// that starts with 4 asks the sender to come back.
if e.Decision == store.SpamDecisionReject && strings.HasPrefix(reason, "4") {
e.Decision = store.SpamDecisionTempfail
}
if err := j.st.InsertInboundSpam(e); err != nil {
log.Printf("log-tailer: inbound journal: record %s -> %s: %v", from, to, err)
}
}
func (j *inboundJournal) remember(queueID string) {
if _, ok := j.port25[queueID]; ok {
return
}
if len(j.order) >= inboundQueueIDs {
delete(j.port25, j.order[0])
j.order = j.order[1:]
}
j.port25[queueID] = struct{}{}
j.order = append(j.order, queueID)
}
// printable cuts s to max bytes and replaces everything outside printable
// ASCII.
func printable(s string, max int) string {
if len(s) > max {
s = s[:max]
}
b := []byte(s)
for i, c := range b {
if c < 0x20 || c > 0x7e {
b[i] = '?'
}
}
return string(b)
}
// lineTime reads the time at the head of a mail.log line and returns it with
// the rest of the line. A refusal is dated by its line and not by when the
// tailer got to it: after the panel was down the two are hours apart. The
// syslog format carries neither year nor zone - the year is taken from now
// (a stamp that would then lie in the future is last year's), the zone is the
// container's, which is what postlogd wrote it in. A line with no readable
// stamp is dated now.
func lineTime(line string, now time.Time) (time.Time, string) {
if m := isoStampRe.FindStringSubmatch(line); m != nil {
rest := strings.TrimSpace(line[len(m[0]):])
if t, err := time.Parse(time.RFC3339Nano, strings.TrimSpace(m[0])); err == nil {
return t.UTC(), rest
}
return now, rest
}
if m := syslogStampRe.FindStringSubmatch(line); m != nil {
rest := strings.TrimSpace(line[len(m[0]):])
stamp := strings.Join(strings.Fields(m[1]), " ")
t, err := time.ParseInLocation("Jan 2 15:04:05", stamp, time.Local)
if err != nil {
return now, rest
}
t = t.AddDate(now.In(time.Local).Year(), 0, 0)
if t.After(now.Add(24 * time.Hour)) {
t = t.AddDate(-1, 0, 0)
}
return t.UTC(), rest
}
return now, line
}
// The journal keeps at most this many rows whatever their age: refusals are
// written at the rate strangers knock on port 25.
const inboundSpamLogMaxRows = 10000
// inboundRetentionLoop prunes the inbound journal at once and then on the
// cadence of the send log.
func inboundRetentionLoop(ctx context.Context, in *Inbound) {
prune := func() {
days := in.Retention()
if days <= 0 {
days = store.InboundSpamLogRetentionDaysDefault
}
n, err := in.Store.PruneInboundSpam(time.Now().UTC().AddDate(0, 0, -days), inboundSpamLogMaxRows)
if err != nil {
log.Printf("log-tailer: inbound journal prune: %v", err)
return
}
if n > 0 {
log.Printf("log-tailer: pruned %d inbound journal rows (older than %d days or beyond %d rows)", n, days, inboundSpamLogMaxRows)
}
}
prune()
t := time.NewTicker(retentionInterval)
defer t.Stop()
for {
select {
case <-ctx.Done():
return
case <-t.C:
prune()
}
}
}