a3050afb667d31568df7b345a40c3362378d7728
132
Commits
| Author | SHA1 | Message | Date | |
|---|---|---|---|---|
  mixandClaude Opus 5.5
|
a3050afb66 |
base-trixie: findings of the reviews
test / test (push) Canceled after 0s
Security review ( |
||
|
mixandClaude Opus 5.5
|
1c6e04b7d3 |
base-trixie: the image on Debian trixie, Postfix 3.10.13; Received: hidden on 465/587
test / test (push) Canceled after 0s
The runtime base is debian:trixie-slim, the build stage golang:1.26-trixie:
Postfix 3.10.13 with upstream's 2026 security fixes up to the August
batch, libmilter 8.18.1, OpenSSL 3.5. Every citation of another program in
code, scripts and documents was re-read against the trixie sources
(dev/upstream; bookworm's kept in dev/upstream-bookworm) and updated;
development.md's cited-package table holds the trixie versions.
What the re-reading changed in behaviour:
- logtail: reject lines with sasl_method/sasl_username/sasl_sender
(smtpd_check.c:1019-1026); cleanup's "removed (canceled)" forgets a
port-25 id (cleanup_api.c:363-364).
- internal/postfix/retry.go reads postconf's stdout only: since 3.9 a
deprecated-parameter warning on stderr made one value too many and the
panel fell back to the compiled-in retry policy.
- logrotate-loop.sh: logrotate 3.22 exits 1 when it ignores a config;
the loop refuses on that output whatever the status.
- postfix-config.sh: cyrus_sasl_config_path set explicitly (bookworm's
07_sasl_config.diff is gone); maillog_file_permissions=0640; mynetworks
explicit; Dockerfile: SYNC_CHROOT= in /etc/default/postfix,
opendkim-tools dropped, openssl named, systemd-standalone-sysusers so
cron does not bring the whole of systemd.
- The selfpost group is 10998, outside the system range 100-999: in
trixie a package took 998 and the build failed (owner's decision:
a conflict-free id, the distribution's own ids untouched). The
entrypoint hands /data back by name when the user or the group differs.
Guide, backup refusal text, help (design commit
|
||
|
mixandClaude Opus 5.5
|
6023feb405 |
audit 6: documents, comments, dead code and staticcheck; the test gaps closed
test / test (push) Canceled after 0s
The last stage of the 2026-10-12 audit; no cut. Doc drift: architecture.md (the sender check in § OpenDKIM, the limit: row's 48 hours, the inbound log and filter-list routes and who may open them, migrations to 0011 and every table, KitPage as production data, Reload configuration as a restore writes, the panel's roles counted), schema-migrations.md (tables and columns since v2), guide.md (a delivery's lines go at the next prune; INBOUND_JOURNAL_MILTER_SOCKET moved to architecture.md's internal list), development.md (the env guard, the deferral the e2e proves, its fixed sleeps), roadmap.md. Comments say what the code does now; TestBaselineIsTheOnlyMigration is TestChainStartsAtTheBaseline; the milter package doc names its three milters. Dead code gone: view.Facts, dmarc HostedSuggestion and ValidateHostedAddress, app UpdateMode (each with the test that alone kept it), the method checks the method-bound routes made dead; Settings.WithResult calls resultFlash. staticcheck 2026.1 in test.yml on both modules, clean: the e2e package builds without its tests now (a test and two constants moved out of non-test files), three findings fixed. The test moved, TestImageBuildPreservesLogrotateMode, had sat in a non-test file and never run; run, it failed on the entrypoint (docker run without a hostname) and now asks with --entrypoint stat. The test gaps no step named, each closed: follow() with mail.log gone past blindMin and made again; CheckRestore with a manifest that is not JSON; the origin check with a port; every /server/ pattern in web.go through server() and in the routes table (it found GET /server/health/fragment missing from the table - guarded, not tested); L9 as the owner decided it on 2026-10-12, the comment on senderSession.Headers and the test passing an h that holds the earlier messages' From:. Plan stage 6 ticked; roadmap audit row done, 81/81. Tested on the test server: go vet + go test of 27 packages, make e2e (all pass, 513 s) on this tree; staticcheck locally on both modules. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com> |
||
|
mixandClaude Opus 5.5
|
a39eda5a0b |
audit 5: findings of the two reviews
- The e2e's numeric-settings test ran a container that passed every gate
and never ended: the good values are checked beside a bad one after
them; each run is named, bounded to a minute and removed.
- The gates refuse a value under the parameter's minimum (a zero window
kills anvil, a zero milter timeout smtpd) and one past an int, in
seconds for a time value (src/global/conv_time.c, src/anvil/anvil.c).
- OpenDKIM waits, five seconds at most, for the relay's socket.
- The relay marks its lines `syslog: `, writes control characters out,
reads into a queue at once and drops past 200 lines a second, saying
how many (security F1, F2; a stuck relay would hold OpenDKIM in
syslog(3)).
- smtp_tls_mandatory_protocols >=TLSv1.2 too.
- extract-cert.sh keeps the old files' group and sets its trap before
mktemp.
- crashexit spares cert-reload; architecture.md "four times running" and
what else exits the container; security.md the restart after five panel
exits and the /dev/log socket.
- Measured against the technical review: smtpd has the selfpost group
(Groups 104 998), so it did reach the supervisor socket before this
stage.
Accepted, reasons in the plan: the logrotate state file in the postfix
user's directory, envdoc_test's reading of `${VAR:-}` only, crashexit's
window pruning untested.
Tested on the test server: go vet + go test of 27 packages, make e2e (all
pass, 592 s) on this tree.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
|
||
|
mixandClaude Opus 5.5
|
dab7715e80 |
audit 5: the supervisor socket the panel's alone, OpenDKIM logs and oversigns From
test / test (push) Canceled after 0s
M17: the supervisord control socket is root:supervisor 0770, a group whose one member is the panel; OpenDKIM and Postfix keep `selfpost` for the files under /data and no longer reach the socket. L30, read in OpenDKIM 2.11.0~beta2's source first: it logs through syslog(3) only and has no Logfile option, and nothing in the container served /dev/log. internal/syslogrelay (the panel binary with -syslog-relay=/run/syslog/log, program syslog-relay before OpenDKIM, /dev/log a link to its socket) passes the datagrams to stdout, error and worse marked. opendkim.conf: Syslog yes, Canonicalization relaxed/simple, OversignHeaders From (RFC 6376 § 3.4, § 8.15). e2e: OpenDKIM's start line reaches the container log; every program RUNNING; DKIM verifies. Tested on the test server: go vet + go test of 26 packages, make e2e (all pass, 626 s) on this tree. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com> |
||
|
mixandClaude Opus 5.5
|
eaa7de0754 |
audit 4: findings of the two reviews
test / test (push) Canceled after 0s
- Two inbound domains on one upstream could never change its TLS mode,
each save refused for the other (medium): SetTransportAs changes the
mode for every domain on the upstream when the one saving may change
them all; a domain user who may not is refused with the mode, not the
other tenant's domain name (security F1: the first version named it).
- A deletion's leftover DKIM key is removed by Resync (Reload
configuration, a restore's first start) for names no domain has.
- The backup's write deadline moves before Check.
- hasUnsavedEdit skips password and file fields and compares a multiple
select's options; htmx.config.timeout 30 s, so a refresh never answered
ends and is tried again; a failed refresh's promise is handled.
- An empty filtered Outbound log says no message matches the filters.
- The dummy bcrypt hash is made at start; a DMARC part over 8 MiB is
refused for its size like the other two limits.
- Health's Configuration box says what Reload writes (design first,
|
||
|
mixandClaude Opus 5.5
|
006bed7022 |
audit 4: DNS outages unknown, DMARC limits said, typed Health, one TLS mode per upstream
test / test (push) Canceled after 0s
M14, M15: queryEach gives each resolver 2 s within the check's context and asks the one that answered last first; an address or PTR lookup not answered is Unknown, an answered mismatch beside it still an error; Checker.Server keeps the last answer's addresses, so SPF does not go Unknown with it. Notifications report faults only, now tested with SPF unknown alone and beside a wrong DKIM. DMARC reports: tests of malformed XML, gzip and zip bombs, a message over the limit, the service's Resync, hosted suggestion and validation. The finding: both limits cut silently and an oversized report failed as damaged; readAtMost refuses it for its size (MaxMessage 10 MiB, moved into IngestMessage). healthReadings replaces map[string]any for Health and Overview. L5: an inbound upstream has one TLS mode - a second domain to the same [host]:port with another mode is refused, naming the first; the TLS policy map holds each next hop once. Tested on the test server: go vet + go test of 25 packages, make e2e (all pass, 631 s) on this tree. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com> |
||
|
mixandClaude Opus 5.5
|
eee089ad4b |
audit 4: cards ordered by number, a notification owed to each recipient
test / test (push) Canceled after 0s
The cards layout (owner, 2026-10-13; the mockup first,
|
||
|
mixandClaude Opus 5.5
|
2b8e0c086d |
audit 4: a deleted domain says so, the Outbound log filters by status and day
test / test (push) Canceled after 0s
L17: past DeleteDomain both rebuilds run, their errors joined under domain.ErrDeletedIncomplete; the key is removed only when OpenDKIM's tables were rebuilt (they may still name it otherwise); the domain list says the domain is deleted and that Reload configuration finishes it, not "internal error". Tests in domain (the tables failing, the sender map failing) and handlers. The Outbound log's status and day filters (owner, 2026-10-13; the mockup first, |
||
|
mixandClaude Opus 5.5
|
ca876c1631 |
audit 4: log paging at its edges, NaN refused, a red rate limit, recalculation asks, reload writes all
test / test (push) Canceled after 0s
Step 4 of stage 4 (L10, L11, L13, L18, L19 and a row of
review-2026-08-followups):
- The Outbound log counts first and clamps the page to the last; a number
too large for an int (strconv's ErrRange with the largest int) is the
last page on both logs. Tested at the edges: first, last, past the last
however far, nonsense, an empty log asked for page 5.
- checkAutoMultiplier is "not within", so NaN and the infinities are out.
- rate-limited is red (components.go; the mockup first,
|
||
|
mixandClaude Opus 5.5
|
07be0cfb58 |
audit 4: sign-in timing and sessions, server timeouts, polling that recovers
test / test (push) Canceled after 0s
Steps 1-3 of stage 4, tested together (their docs share files).
Step 1 (M8, L15, L16): an unknown name pays a bcrypt comparison at the
panel's work factor too (noSuchUserHash, made once) -
TestLoginTakesAsLongForAnUnknownName holds the medians within a factor of
two; the Account form's current-password check has its own limiter, per
user (AllowPasswordChange); a sign-in destroys every session the request
presents before it issues its own.
Step 2 (L14): newHTTPServer - ReadTimeout 30 s, WriteTimeout 60 s,
IdleTimeout 2 min beside ReadHeaderTimeout; the full backup moves its write
deadline to two hours through http.ResponseController (e2e: the log shows
no failure to move it).
Step 3 (M11), read in htmx 2.0.4's source (dev/upstream/htmx-2.0.4, the npm
build byte-equal to the shipped htmx.min.js): the poll passes source: box,
and the next refresh is scheduled from htmx:afterRequest, which htmx fires
on load, error, abort and timeout, and after a swap on the nearest parent
left (the box found by detail.target's id); failures in a row double the
wait to 60 s. RequireAuth answers htmx with 401 + HX-Redirect: /login,
which htmx follows before it looks at the status. testdata/poll_test.js
runs panel.js in node against a stand-in DOM and clock (TestPanelJS; skipped
without node, which the test server lacks - run here, and it fails on the
old panel.js).
Also measured: the stage-3 e2e regression for H1 fails as it should on HEAD
with the old entrypoint ("check data/postfix/e2e_extra_map: permission
denied", the panel restarting).
Tested on the test server: go vet + go test of 25 packages, make e2e (all
pass, 639 s) on this tree.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
|
||
|
mixandClaude Opus 5.5
|
242515c305 |
audit 3: backup CLI, envelope budget, directory sync, docs; findings of the reviews
Steps 3-6 of stage 3 and what the two reviews found in steps 1-6, in one commit: the review fixes sit in the same hunks. Step 3, CLI backup (M3, M5, L1, L2; L4 of the stage's list): - backup.snapshotDB refuses a database that is not there and opens the source mode=ro (measured: mode=ro creates no file). - selfpost-backup holds an encrypting password to validate.MinSecretFilePasswordLen; -o of both directions writes a temporary file beside the name and renames it whole; -decrypt peeks the first chunk (password and type authenticated) before the output exists. - what a backup leaves out (a link under data/, a link under certs/ to a directory, a socket...) is told to OnWarn; certs/ with no file is backup.ErrNoCerts, refused by Check and Create, named by the panel. - parseTrustedProxies: a bare address takes /128 when written with a colon (::ffff:10.0.0.1/32 was ::/32). Step 4, secretfile (M4, L3): scrypt's V at most 128 MiB, r <= 64, p <= 2; bytes after the last chunk are ErrCorrupt. Step 5 (L7): atomicfile.SyncDir after the rename, used by Write, replaceSASLDB and the CLI; atomicfile_test.go. Step 6: security.md and guide on backup and restore. Reviews (technical, security): - H1: a restore of a server with DMARC on was refused on every start - the archive records no owner, tar as root left the DMARC maps root's, and the entrypoint only chowned the maps it named. It now hands every file beside queue/ in /data/postfix to panel:selfpost 0640; e2e writes an extra map before the backup and finds it panel:selfpost 640 after a root unpack. - the comparison reads regular files only (Lstat, size, LimitReader); a record unpacked again wins over the earlier check's mark; ConsumeRestore removes both; the mark is synced; a journal file beside a freshly unpacked database is refused. - AttachRefusalLine refuses an empty line as the old statement did; skipped entries named in words; -decrypt of an empty payload is no failure; guide wording on the shell's '>' and a killed CLI; the in-place restore's 'mv .' (older than this stage) fixed. Tested on the test server: go vet + go test of 25 packages, make e2e (all pass, 643 s) on this tree. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com> |
||
|
mixandClaude Opus 5.5
|
508ac96c79 |
audit 3: every per-message question of the send log has its index
test / test (push) Canceled after 0s
Migration 0010 (DDL only; 0009 went to stage 2b). EXPLAIN QUERY PLAN of
every send_log statement showed, beyond M2 as written (stage 2 had already
moved the by-domain counts onto (domain_id, created_at)):
- the tailer's sweep of open rows (every 5 min) and its count scanned the
table: partial idx_send_log_open (created_at) WHERE status IN ('queued',
'deferred'), the statuses literal in the statements, ORDER BY +id (else
SQLite walks the table in id order to skip a sort);
- AttachRefusalLine's NOT EXISTS (... refusal_line = ?1) scanned the table
for every refusal line: partial idx_send_log_refusal_line, the query
carrying the index's WHERE; its row found by idx_send_log_rejected
(status literal, +queue_id) instead of every refusal without a queue id;
- an application's level-2 count went through its domain's index:
+domain_id in appByLogin;
- a page of one domain's or one application's log sorted all of it:
idx_send_log_domain_id, idx_send_log_app_login (id order; several domains
still sort, top-N). idx_send_log_domain stays for the global filter.
The statements are constants and builders (countMessagesQuery,
peakPerHourQuery, listOpenQuery, countOpenQuery, sendLogPageQuery,
attachRefusalLineQuery, limitRefusalsQuery) and
TestSendLogQueriesUseTheirIndexes explains exactly those: the index each
must use, no SCAN of send_log, no sort where none is wanted; the literal
statuses checked against the constants. It replaces the hand-written plan
check of TestLimitRefusalsSince.
Tested on the test server: go vet + go test of 24 packages, make e2e (all
pass) on this tree.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
|
||
|
mixandClaude Opus 5.5
|
8e3241ef98 |
audit 3: a restore is ended by its Resync, and half an archive is refused
test / test (push) Canceled after 0s
The restore manifest was consumed as soon as its version was checked, before store.Open and the Resync; a start that failed in between left a data directory the next start took for an ordinary one (M1). openData now runs CheckRestore, store.Open, the Resync, and only then ConsumeRestore. A truncated archive had nothing to refuse it: the manifest is the first entry, so a download cut short or a tar stopped by a damaged block was a restore on half a database. Every archive now ends with data/contents.json, the size and SHA-256 of each data/ file (deploy files are the operator's to edit), and CheckRestore compares /data with it before the database is opened, then renames it contents.checked.json so a start after a failed one does not compare again unless the manifest changed. Tests: a start whose Resync fails, then a start that resyncs and consumes; an archive cut in the database, before the deploy files and in its last entry (refused, database untouched, no -wal); a damaged key named; the record's names confined to data/; the domain import of a cut plain export, a cut and an altered .spde (refused, nothing created). e2e on the test server: full_backup_restore now also sees contents.json in the archive and gone after the first start, through the real entrypoint and daemons (all pass, 762 s); go vet + go test of 24 packages pass there. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com> |
||
|
mixandClaude Opus 5.5
|
9eddd49602 |
audit 2c: findings of the two reviews
test / test (push) Canceled after 0s
Technical review (Opus, 11 findings, 2 medium) and security review (Fable, no high or medium) of 3c86280..HEAD. Both confirm the port gives dkim_mail_parse()'s result on every input it parses, and that no line the check passes is signed with another domain's key. - F1 (medium): OpenDKIM's parser takes lines no mail program reads as one address - "ceo@bank.com:me@mine.com", "Name <ceo@bank.com:me@mine.com>", "<b@evil.com, a@mine.com>", "ceo@bank.com <me@mine.com>", "ceo@bank.com@mine.com" - and signs for the last while a lenient reader shows the first. The line must now also be one address in RFC 5322's grammar (net/mail, display names in any charset passed through), at the domain OpenDKIM signs for. A quoted display name may still hold anything. - F2 (medium): the bindings were asked about OpenDKIM's form of the address, quotes and all, so "x"@mine.com passed for an application holding @mine.com while x@mine.com belonged to another. They are now asked about the reader's address, its local part unquoted, as Postfix holds the envelope's. - F3 / security (low): Bindings.Allowed and DomainOf fold case in ASCII only: Postfix folds as Unicode (casefold.c, ICU), but every key and login the panel writes is ASCII, and the Unicode fold let the Kelvin sign pass for "k" (net/mail then refuses such a line on the domain test). - F4 (low): a TERM between "postfix start-fg &" and "child=$!" is no longer lost: the wrapper stops Postfix as soon as the pid is known. - F5 (low): for startsecs after a start any failed reload passed as "not running". postfix-reload is now build/postfix-reload.sh: exit 75 when master -t says no master runs, else exec postfix reload; RunOnce returns EndError with the status, and only 75 is taken as done. supervisor.State goes. - F6 (low): crashexit spares postfix-reload too - a reload that could not be spawned is reported to whoever asked, not a reason to stop the container. - F7 (low): a certs/ that is itself a link is resolved before the walk (filepath.WalkDir does not follow its root). - F8 (low): at shutdown the tailer reads what is left of the rotated file before saving its place. Not changed: a refusal line still waiting on the old file at shutdown is not kept - the panel writes refused rows itself, only a kill loses one, and that window is the minute's grace. - F9 (low): pct read as text - empty or unreadable is 100, the rest held to 0..100. - F10 (low): the directory is synced after sasldb2's rename. - F11 (low): POSTFIX_STARTSECS by expr, so "010" or "08" in MILTER_WAIT_TIMEOUT are decimal, not octal or an error. - Security (low): pagehide also empties defaultValue - the input's value attribute and the textarea's text - not only .value. A sasldb2.new left by a killed panel is removed at start (entrypoint) and never archived. security.md says that links under certs/ are followed and by whom they can be placed. Tests: sender check cases for F1 (four lines), F2 (a bound address quoted, for its owner and for another), F3; EndError's state and status; pct cases. architecture.md and CHANGELOG describe the reader rule. Stand (2026-10-15, make e2e, exit 0); go vet and go test ./... pass on the server (24 packages). Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com> |
||
|
mixandClaude Opus 5.5
|
ec08c8606b |
audit 2c: the From: parser port checked against OpenDKIM's signing itself
test / test (push) Canceled after 0s
The last check of the port, and the plan's last line of stage 2c. Through
Postfix the sender check refuses most lines before OpenDKIM sees them, so
TestPortAgainstOpenDKIM (internal/milter) talks to OpenDKIM directly:
test/upstream/opendkim-probe.sh starts OpenDKIM from the SelfPost image in
a throwaway container, with keys and a refile SigningTable for three
names, and the test plays it each From: line through go-milter's client,
as a client OpenDKIM trusts (127.0.0.1, InternalHosts' default). For each
line the d= of the signature OpenDKIM adds must be the domain the port
predicts - its address, the domain cut to DKIM_MAXHOSTNAMELEN, the whole
cut to MAXADDRESS, matched against "^.*@<domain>$" case-insensitive as
dkimf_apply_signtable() matches - or neither signs.
Run on the test server, 2026-10-15 (OpenDKIM Filter v2.11.0 from the
image): 385 lines - both shared files and lines composed from display
names plain, quoted, escaped and encoded, comments nested and open,
groups, lists, brackets, source routes, domain literals and quoted local
parts, with the other sender's name and the server's in each place - 385
agree, 287 signed. The two lines padded so that, cut at 256 bytes, they
end in "@other.e2e.test" are signed by OpenDKIM with other.e2e.test's key:
the danger the sender check's 256-byte rule (
|
||
|
mixandClaude Opus 5.5
|
3584adcbc2 |
sender check: From: read with OpenDKIM's own parser; L8 closed
test / test (push) Canceled after 0s
Audit 2026-10-12, stage 2c, the owner's decision of 2026-10-14: the check
decides by the address OpenDKIM signs for, whatever a display name or a
comment holds.
internal/milter/mailparse.go ports libopendkim's dkim_mail_parse() with
dkim_mail_first_special(), dkim_mail_token() and
dkim_mail_matching_paren() from Debian's opendkim 2.11.0~beta2-8+deb12u1,
line for line: the '<' form, group and list punctuation, an '@' before the
first ':', ';' or ',' deciding where the address starts, source routes
dropped, quoted strings and domain literals kept with their delimiters,
comments skipped, whitespace between atoms one space, no encoded word
decoded; and the C's own quirks ("@d" gives d as the local part too).
testdata/mailparse_harness.c compiles the C from the image's source; the
test feeds it and the port 200 019 lines (the hand-written cases and
200 000 random ones over the specials) - 93 443 parse and agree byte for
byte, 106 576 fail and agree on the error (on an error the C leaves its
in-place result unterminated, and OpenDKIM uses neither part then).
judge(): one From: field and one address stay required (moreAfter: only
list or group punctuation may follow); what OpenDKIM cannot read is
refused (mlfi_eoh() would not sign it); the address is lower-cased in
ASCII only, as dkimf_lowercase() and Postfix's tables do. Found reading
dkimf_apply_signtable(): OpenDKIM matches its refile SigningTable against
"user@domain" printed into MAXADDRESS + 1 = 257 bytes, so a longer address
is matched cut short - a quoted local part padded to end there in
"@another.tenant" would be signed with that tenant's key while its real
domain is the application's. Such an address (over 256 bytes) is refused;
one that fits is matched whole. The display-name and comment scan, the
"one @ outside quotes" rule and the SigningTable reader (Signers) go; the
check asks for header values with their leading space as OpenDKIM does
(OptHeaderLeadingSpace; Postfix keeps a v2 milter's event mask, milter8.c
milter8_negotiate()).
L8 closed: refused or passed no longer depends on other domains of the
server; security.md's accepted risk is struck. The shared other-sender
lines are now held to "never signed for the other sender": unit test (a
passed line is read at the application's domain) and e2e (a passed line
arrives signed for the application's domain, none for the other).
Documents: architecture "How the header is read", security's sender
check paragraph, the guide; the help topic per
|
||
|
mixandClaude Opus 5.5
|
cad1182150 |
audit 2c: what only the stand could answer, measured and written down
test / test (push) Canceled after 0s
All 2026-10-15; each written beside the code or text it settles. - A-43: /usr/sbin/postlog is -rwxr-xr-x root:root in the image (postfix 3.7.11-0+deb12u1; dpkg-statoverride lists postdrop and postqueue only): Debian does not install it set-gid, so it is no use to dmarc-ingest (cmd/panel/dmarc_ingest.go). - C-X1: OpenSSL 3.0.22 in the image prints the same public key from `x509 -pubkey` and `pkey -pubout` for RSA-2048, P-256, P-384 and a SEC1 P-256 key - cert-reload's same_pair holds for certbot's ECDSA keys. - C-K5, D-guide:933/:233: on the live server (Docker 29.6.1, userland proxy on 0.0.0.0 and ::, no IPv6 on the network) an outside IPv4 client and the host's own public IPv4 address are logged as themselves; the host's IPv6 address and 127.0.0.1 arrive as 172.18.0.1. Over IPv6 every client is the gateway: one level-1 budget, an application IP list cannot name them, rspamd sees the gateway. Guide paragraph under level 1, notes at the milter and allow-list, comments at the anvil setup and authip.go, and a Security entry in the CHANGELOG (an allow-list naming the gateway would admit every IPv6 client). - D-guide:61: a published port nothing listens on (a throwaway container, -p 20999:25) is refused from outside over IPv4 and from the host's public IPv4, accepted and closed through 127.0.0.1 - Docker's proxy, which IPv6 goes through too (not probed over IPv6 for this). - C-M8: the live installation has received one DMARC report, of one record: nothing to set maxReportRecords by yet - said at the constant. - D-S7, D-arch:516, D-arch:532: a new database filled through the store to 120 days at 1 900 messages a day (228 000 messages, 1.37 million lines, 464 MB) on the test server (2 cores): the longest write is the line prune after a 30-day gap, 1.9 s for 342 000 lines (0.4 s with nothing to delete); the send-log delete 0.6 s; 2 000 lines of a tailer pass 18 ms; VACUUM INTO 1.7 s; a second connection writing every 50 ms waited 1.8 s at most, no SQLITE_BUSY - inside busy_timeout's 5 s. architecture.md's "about a second per million lines" was five times too fast; its "six to ten lines of 200 bytes" is 6.0 lines of 146 bytes on the live data. - D-B16: which favicon is fetched, by the probe server's log: Edge 155 the SVG, the desktop app's embedded Chromium 152 the PNG, with or without sizes="any" - layout.html's "vector wins where supported, raster catches the rest" holds; Firefox and Safari not measured. - D-arch:121: the allow/deny trial again, in rspamd/rspamd:3.11.1-3 with today's multimap.conf: every mix of allow and deny entries matching one message ended reject; a changed map was used about 10 s later. - D-sec:424: rspamd runs as _rspamd, uid 11333; the panel is uid 999. Stand: go vet and go test ./... pass on the server (24 packages); make e2e exit 0. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com> |
||
|
mixandClaude Opus 5.5
|
fbd0acd035 |
audit 2c: the rest the sources contradict, and the documents
test / test (push) Canceled after 0s
Code and config: - C-M7: a DMARC report's explicit <pct>0</pct> stays 0; only an absent pct takes RFC 7489 § 6.3's default 100 (Pct *int). - C-D1: docs/development.md now tables the Debian versions the citations were read at (postfix 3.7.11-0+deb12u1, opendkim and libopendkim11 2.11.0~beta2-8+deb12u1, libmilter1.0.1 8.17.1.9-2+deb12u2, libsasl2-2 and sasl2-bin 2.1.28+dfsg-10, supervisor 4.2.5-1, logrotate 3.21.0-1 - read off the image with dpkg-query, 2026-10-15); the e2e stand's new checkCitedPackageVersions fails when the image it built differs. - C-E2: the e2e zone's SOA serial only grows (CoreDNS v1.14.6's file plugin reloads on a grown serial, not on mtime); Corefile comment says so; TestZoneSerialGrows. - C-K8, C-K9: the e2e compose and stage comments say how --project-directory and .env resolve. - D-arch:154: postfix-config.sh's "Postfix insists" comment says what TLS_README advises and that smtpd reads the key as root. - B-G7: inflight.go's reservationTTL comment: a dropped client does give ABORT (smtpd_proto() -> mail_reset() -> milter_abort()); the TTL is for smtpd dying or the milter connection breaking. - D-guide:301: the nginx fragment's first-certificate command now reaches certbot (--entrypoint certbot, --standalone, port 80 free), and the renewal loop passes --webroot -w /var/www/certbot - a standalone certificate would otherwise renew standalone while nginx holds port 80 (certbot docs, "Renewing certificates": the issuing plugin is used "unless you specify other plugins or options", and options of a successful renewal are saved). The register's proposal missed that. Documents: every D-arch, D-guide and D-sec claim the register's Answers E found false or partly true, rewritten - by three Sonnet agents, one per document, then reviewed: two proposals were obsolete after this stage's fixes and not used (a bounce's From: line - extract_from = "smtp" now; "Origin: null" - Referrer-Policy is same-origin now); a stand observation the agent had dated from a commit is now tied to the e2e test that checks it; the SQLite writers paragraph describes _txlock=immediate; the log tailer's rotation paragraph describes the minute's grace; the guide's nginx steps were put in working order (links and modes before `up`, no reconfigure step). "Deferred" is kept for Postfix's queue in the guide too. guide.md's CRLF from the agent normalised. Stand (2026-10-15, make e2e, exit 0, the version gate and zone serial test included); go vet and go test ./... pass on the server (24 packages). Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com> |
||
|
mixandClaude Opus 5.5
|
2e1544dac5 |
certbot: certs/ links into /etc/letsencrypt, mounted at the same path; the backup follows them
test / test (push) Canceled after 0s
C-A1, C-Q4, D-guide:258, :291, :722. certbot's live/<host>/*.pem are relative links into ../../archive/<host>/ (eff-certbot docs, "Where are my certificates?": point a server at them "or create symlinks"). The guide had `ln -s /etc/letsencrypt/live/<host> certs` and the nginx fragment mounted live/<host> alone: inside the container the links led into a directory it did not have - postfix-config.sh's [ -f ] false, no copy, Postfix without a certificate - and the panel's backup, reading certs/ through /selfpost-deploy, skipped links silently. Now certs/fullchain.pem and certs/privkey.pem are links to /etc/letsencrypt/live/<host>/, and /etc/letsencrypt is mounted read-only at the same path: a commented line in deploy/docker-compose.yml for the Apache scenario, mounted in the nginx fragment (./nginx/certbot-etc). The guide's steps add certbot's own commands for a server that is not root - chmod 0755 live archive, chgrp 998 + chmod 0640 the key, which certbot keeps on renewal - so the panel reads them (Health, backup). Postfix's side follows links already ([ -f ], cp, cmp, openssl). The vhost comment says the same. The backup table: links are followed by the panel's backup; a stopped tar stores them as links unless -h (GNU tar). internal/backup: addTree follows, for certs/ only, a link to a regular file and archives that file under the link's name; a link to nothing reachable is an error naming it and the mount. Tests: TestCertsLinksAreFollowed (certs/ -> live/ -> archive/, the archive holds the files; the tree gone, Check names the dangling link; Linux). Stand (2026-10-15, make e2e, exit 0): the stand mounts ./letsencrypt at /etc/letsencrypt; new certbot_layout_followed puts a new pair into letsencrypt/archive/, relative links in live/, absolute links in certs/, recreates the container: 465 serves the new serial, Health names no renewal waiting, and the full backup holds certs/*.pem as regular files with the archive's bytes. go vet and go test ./... pass on the server (24 packages); the Linux-only tests of this session run there and pass. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com> |
||
|
mixandClaude Opus 5.5
|
3c63ecc1ad |
store: migrations under BEGIN IMMEDIATE, the version read again inside
test / test (push) Canceled after 0s
D-S3. A deferred BEGIN starts nothing until the first access (SQLite, lang_transaction.html § 2.2), and migrate() read user_version outside the transaction: after an upgrade the panel and dmarc-ingest, opening the file at once, could both read the old version, both begin, and the second applied the next migration again on the migrated file - "already exists", that Open failing (nothing damaged: its transaction rolled back). The DSN now has _txlock=immediate (modernc.org/sqlite v1.60.1, sqlite.go applyQueryParams() sets beginMode, tx.go newTx() issues "begin immediate" for any transaction not read-only), so a migration holds the write lock from its BEGIN, and migrate() reads user_version again inside it and skips one another process applied meanwhile. Test TestTwoOpensMigrateOnce: a raw connection holds BEGIN IMMEDIATE on a new file while two Opens read version 0 and queue on their first BEGIN; released, both succeed and user_version is the number of migrations. Without the fix it fails as the register said: "apply migration 0001_init.sql: SQL logic error: table settings already exists (1)". Stand (2026-10-15): go vet and go test ./... pass on the server (24 packages); make e2e exit 0. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com> |
||
|
mixandClaude Opus 5.5
|
9ca50d33fd |
web: Referrer-Policy same-origin; a password shown once emptied on pagehide
test / test (push) Canceled after 0s
D-B3. Fetch Standard § 3.2 "append a request `Origin` header": under Referrer-Policy no-referrer a request that is not GET/HEAD and not CORS - a form POST, a navigation - gets Origin "null". originAllowed() refuses "null" whenever Sec-Fetch-Site is absent, and Fetch Metadata § 3 sends Sec-Fetch-Site only to a potentially trustworthy URL: every form, sign-in included, failed in Safari before 16.4 and in any browser on a plain-HTTP panel at a name. same-origin nulls Origin only for a request not same origin with its URL and still sends no referrer to another site. The comments of the origin check say so. D-B13. noStore keeps the page out of HTTP caches (RFC 9111 § 5.2.2.5); a history mechanism "might honor" it (§ 6), and Pragma in a response "was never specified" (§ 5.4) - it goes. CopyField.Secret marks the password of out-app-created data-secret, and panel.js empties such fields on pagehide (HTML § 7.5.9, "unload a Document", step 10.2: fired whether or not the document is kept for history traversal). Measured 2026-10-15 in Chrome 152 (the desktop app's browser), a local probe page with Cache-Control: no-store and a data-secret field: Back reloads it (pageshow persisted=false, a new load), so this browser keeps no-store pages out of its back/forward cache; the handler, fired on the same page, empties the field. Firefox and Safari not measured. Tests: the header value; the created page has the password data-secret and the login not, no Pragma. Stand (2026-10-15, make e2e, exit 0): new form_post_with_the_browsers_origin - Referrer-Policy: same-origin on a page; a POST carrying the page's own Origin and no Sec-Fetch-Site, what a browser sends now, passes; one carrying Origin: null, what it sent before, is 403. (A first run failed where the e2e client and a view test scraped the old input markup; both read data-secret now.) go vet and go test ./... pass on the server (24 packages). Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com> |
||
|
mixandClaude Opus 5.5
|
98e09fef4a |
health: sockets dialled, the certificate Postfix serves, "deferred" only for the queue
test / test (push) Canceled after 0s
Audit 2026-10-12, stage 2c, "Health says what is served"; the pages of
ea8b6bf's mockup.
Sockets (A-38). A unix socket's file outlives a killed OpenDKIM or panel
and refuses every connect() (unix(7)); Health said "Listening" from a
stat. CheckSocket and CheckMilter's unix case now dial: refused is
"Nothing listens on the ... socket: the program that made it has
stopped", EACCES is unknown (the panel may not connect), anything else
"does not answer". The probe closes before a byte, which no milter logs:
libmilter takes the EOF as SMFIC_EOF without a line (sendmail
8.17.1.9-2+deb12u2, the image's libmilter1.0.1, libmilter/comm.c
mi_rd_cmd(), engine.c mi_engine() - its source is now in dev/upstream),
OpenDKIM's mlfi_close() has no connection to free, go-milter v0.4.1
returns quietly on io.EOF (session.go, HandleMilterCommands()). The page
polls every 5 s while in view, so this mattered.
Certificate (A-65). Postfix serves build/postfix-config.sh's copy,
refreshed by cert-reload every TLS_RELOAD_INTERVAL_SECONDS. The panel now
reads smtpd_tls_cert_file from postconf -h at its start (as the retry
policy) and TLS_RELOAD_INTERVAL_SECONDS; CheckServedCertificate checks the
served copy and, when the mounted file differs byte for byte (cert-reload's
own cmp -s), names it as Renewal: its expiry and mtime, two facts on the
card. Waiting is fine for the interval: a nearing expiry of the served one
is then no warning (an expiry still an error); past interval + 1 h it is a
warning pointing at cert-reload's lines. The notifier sends no expiry
notice while a renewal waits on time.
Wording (D-P1, D-P3, and the same for the spam filter): Sender check
note, help topics sender-check, upstream, filter, certificate - word for
word as in the mockup; the filter row's detail too.
Tests: stale socket file (Linux) refused for CheckSocket and CheckMilter;
CheckServedCertificate (same, waiting, late, expired + waiting, no served
path); intervalText; the card's facts; the notifier quiet while a renewal
waits, speaking once it is late. Stand (2026-10-15, make e2e, exit 0):
new health_shows_what_is_served - the three sockets dialled ok by the
panel's user; a certificate mounted without its key is refused by
cert-reload ("do not belong together") and the card shows the served
expiry, "Renewal expires" with the mounted one's, "Renewal mounted", and
"Postfix still serves the certificate above."; a matching pair is then
served again. go vet and go test ./... pass on the server (24 packages).
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
|
||
|
mixandClaude Opus 5.5
|
d20009c0f0 |
sasldb2 changed through a copy renamed over it, never in place
test / test (push) Canceled after 0s
Cyrus SASL 2.1.28 opens sasldb2 with no Berkeley DB environment, so with
no locks (sasldb/db_berkeley.c, berkeleydb_open(): db_create(mbdb, NULL,
0)); saslpasswd2 puts and deletes in the file itself (_sasldb_putdata()),
while smtpd opens and closes it at every AUTH (_sasldb_getdata()) and the
backup copies it whenever it runs. Either could read a half-written page:
a spurious 535, or a torn sasldb2 in a backup whose restore breaks every
application's AUTH (assumptions C-S3, C-Q3).
SASLDB.Set and Delete now run saslpasswd2 on sasldb2.new - a byte copy of
the database, or no file when there is none yet (saslpasswd2 creates it,
DB_CREATE) - then chmod 0640, fsync, rename over sasldb2; a failed run
removes the copy and leaves the database; a copy a killed panel left is
removed first. A mutex keeps two changes of the panel from copying the
same old file. The backup needs no lock after all (the plan said one): it
reads each file through one descriptor, and a rename leaves that file
whole; the walk now leaves out a file that vanished between listing and
open (a writer's temporary file), while the SQLite snapshot's absence
stays an error. CHANGELOG also gains the entries of
|
||
|
mixandClaude Opus 5.5
|
f2ad699473 |
supervision as supervisord does it: Postfix stopped, give-ups FATAL, reload's exit read
test / test (push) Canceled after 0s
C-P42. postfix-script's start-fg runs master as its child and master calls setsid() (Postfix 3.7.11, conf/postfix-script; src/master/master.c); supervisord signals the pid alone (supervisor 4.2.5, process.py, Subprocess.kill()). A stop of program postfix left master running - a restart then found "already running" - and a container stop cut Postfix off with SIGKILL. The wrapper now stays the parent of "postfix start-fg" and on TERM/INT runs "postfix stop", asking again while start-fg is still in its checks (master -t). stopwaitsecs=15. C-V6. The wrapper's give-up on the milter sockets came after the default startsecs, so it was EXITED and restarted for ever. startsecs is now MILTER_WAIT_TIMEOUT + 10, exported by the entrypoint as POSTFIX_STARTSECS and expanded by supervisord (options.py, processes_from_section(); a non-number MILTER_WAIT_TIMEOUT stops the start): a give-up is "Exited too quickly", BACKOFF, FATAL after startretries, crashexit. For those seconds after each start the program is STARTING (the healthcheck's 90 s start period covers it; the sender check counts no downtime then). C-V5. logrotate had startsecs=0, so a refused config restarted it every second for ever. startsecs=1: FATAL after startretries; crashexit reads the event's processname and leaves logrotate's FATAL on Server > Health. B-U2. "supervisorctl start" returns when a startsecs=0 program is RUNNING, not when it ends, and mapped "already started" to success. The panel now speaks supervisord's XML-RPC itself (internal/supervisor, RunOnce): startProcess without waiting, getProcessInfo until it leaves STARTING/RUNNING, exitstatus 0 or an error; a run in progress is waited out and one of its own follows. A failed reload while program postfix is not RUNNING is logged, not an error: the next master reads the files. C-V10: postfix-reload rests EXITED after a run; the e2e process check takes STOPPED or EXITED. C-P34: the dead "postfix set-permissions" branch goes - postfix check's create-missing builds the queue tree. Stand (2026-10-14, make e2e, exit 0; MILTER_WAIT_TIMEOUT=10 on the stand): new supervision_stops_postfix_and_gives_up - supervisorctl restart postfix writes postfix-script's "stopping the Postfix mail system" and leaves one master; with the panel's sockets removed, start answers "postfix: ERROR (spawn error)" (finish() sets spawnerr "Exited too quickly", onwait() reads it first) and Postfix comes up once the panel is back; a group-writable logrotate config ends FATAL with the container up; the panel's Reload configuration answers only after mail.log has "refreshing the Postfix mail system", postfix-reload EXITED. session_survives_restart sees the stop line for the container's restart. Measured: master's own "terminating on signal 15" never reaches mail.log (master TERMs its group, postlogd included, before it logs). go vet and go test ./... pass on the server (24 packages; new unit tests of RunOnce against a fake supervisord). Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com> |
||
|
mixandClaude Opus 5.5
|
463d4ff3fe |
log tailer: the rotated mail.log read until postlogd lets it go
test / test (push) Canceled after 0s
postlogd opens maillog_file once per process (Postfix 3.7.11,
src/postlogd/postlogd.c, pre_jail_init()) and leaves it only when it
exits: at "postfix reload" master closes the service's status pipe
(src/master/master_service.c, master_restart_service() ->
master_status_cleanup()) and dgram_server_abort() exits on reading its end
(src/master/dgram_server.c). logrotate renames and creates, then runs the
reload (build/logrotate-mail.conf postrotate; logrotate 3.21.0,
logrotate.c, rotateLogSet(), after rotateSingleLog()), so lines go to the
renamed file in between, and the tailer, already on the new one, never
read them (assumptions A-33). follow() now keeps the renamed file open as a
rotatedTail and reads it on each tick for rotatedGrace (a minute, far
longer than the gap); its places are never saved as the resume point, so a
restart still reads the rotated file through readRotated. A truncated file
is not kept.
The RFC 3339 stamp goes (A-12): it was written for a maillog_file_format
said to arrive in Postfix 3.9; there is none in 3.7.11 nor in the 3.12
snapshot of 2026-09-25 (proto/postconf.proto; msg_logger.c,
msg_logger_print() keeps strftime("%b %d %H:%M:%S ")). Every fixture now
carries the stamp Postfix writes; logtail's TestMain fixes time.Local to
UTC, since the stamp is local time and this PC is UTC+3.
Tests: TestFollowReadsTheRotatedFileAfterTheSwitch (Linux: a line appended
to the renamed file after the switch reaches the store). Stand
(2026-10-14): go vet and go test ./... pass on the server (23 packages);
make e2e exit 0, delivery_log_survives_rotation included. The tested tree
was this commit plus the previous one; the CHANGELOG entry was written
after the run, and a comment's logrotate citation corrected
(rotateLogSet runs postrotate, not rotateSingleLog).
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
|
||
|
mixandClaude Opus 5.5
|
c8f669714e |
journal milter: recipients as Postfix logs them, once each
test / test (push) Canceled after 0s
The row's recipient has to be the address Postfix writes in to=<..> or orig_to=<..>, or the delivery line finds no row and the sweep closes it as bounced. Postfix 3.7.11 hands the milter the RCPT argument as the client wrote it (smtpd.c, milter_argv(); extract_addr() leaves it), go-milter v0.4.1 trims "<>" and drops the ESMTP parameters (session.go, case CodeRcpt) - so cleanAddress's cut at the first blank broke a quoted local part (B-G5). Postfix keeps the address in internal form (tok822: quotes off, backslash escapes taken) and logs it external (info_log_addr_form.c, quote_822_local(): quoted again only when not a dot-string); recipientAsLogged does the same, "ab"@x -> ab@x, "a b"@x stays. A recipient given twice in any case is one row: cleanup drops the duplicate, comparing folded (cleanup_out_recipient.c, been_here(), BH_FLAG_FOLD; A-31). A rewritten recipient is matched by orig_to=, which log_adhoc() writes only when it differs (A-9). The store's case-insensitive match stays, now with its reason from the source (D-S20). Stand (2026-10-14, make e2e, exit 0): new recipients_as_postfix_logs_them - RCPT TO "x y"@, "ab"@ and "X Y"@ sink.e2e.test give two rows, "x y"@sink.e2e.test and ab@sink.e2e.test, both sent. go vet and go test ./... pass on the server; a unit test drives the milter through go-milter's own client. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com> |
||
|
mixandClaude Opus 5.5
|
0ed03b2094 |
log tailer: expiry, Postfix's time values and refusal lines as Postfix writes them
test / test (push) Canceled after 0s
Expiry (assumptions A-8). Postfix 3.7.11 logs the end of a message's queue lifetime once per message, "<id>: from=<..>, status=expired, returned to sender" (src/qmgr/qmgr_active.c, qmgr_active_done_25_generic()), with no line per recipient; the tailer waited for "to=..., status=expired", which no Postfix writes, and the sweep closed only queued rows, so such rows stayed deferred. The tailer now reads qmgr's line (ExpireMessage: each recipient's newest open row of the id becomes bounced); the sweep closes a deferred row too when Postfix no longer holds the message, and counts as missed only a final line (sent, bounced, expiry). The queue listing's '#' (forced expiry, showq_compat.c showq_message()) is read as a flag (B-PQ1). Time values (B-PQ10, B-PQ11): conv_time() ported - one number, one lower-case unit, the parameter's default unit when none (get_def_time_unit: d for the lifetimes, h for delay_warning_time), the rest ignored; postconf -h -x expands $name; bounce_queue_lifetime capped at maximal_queue_lifetime as qmgr_post_init() caps it, so LongestLifetime is the queue lifetime. Refusals (A-21, A-23, A-24, A-15): smtpd_whatsup() ends the NOQUEUE line with sasl_username=<login> on an authenticated session; the line is now matched to its row by client, login and sender, and a line without a login is no application's. The client= line's login ends at the comma. The "IPv6:" trim goes (namaddr is built from state->addr). A row recovered from cleanup's line says that line names the last recipient only. Stand (2026-10-14, make e2e, exit 0): new expired_message_is_closed - defer_transports=smtp makes the retry service defer the message, postsuper -e and postqueue -i force its expiry; mail.log has "<id>: from=<alerts@sender.e2e.test>, status=force-expired, returned to sender" and the row goes deferred -> bounced; refusals_show_postfixs_line passes with the login match. go vet and go test ./... pass on the server. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com> |
||
|
mixandClaude Opus 5.5
|
55aef6e721 |
sender check: the sender map read and looked up as Postfix does it
Postfix 3.7.11 takes the owners of the first key mail_addr_find() finds -
the address, then without its extension, then @domain, each tried only
when the one before is absent (src/global/mail_addr_find.c,
mail_addr_find_opt()) - and compares the login with them by
strcasecmp_utf8() (src/smtpd/smtpd_check.c,
reject_auth_sender_login_mismatch()). The check accepted a From: address
when any of the three keys named the application: an application holding
@domain could write an address listed for another application of the
domain, which Postfix refuses it in the envelope, and have it signed
(assumptions B-PS1).
The table is now read as dict_thash_open() reads it: continuation lines,
owners split on commas and blanks (B-PS4), the key ending at the first
blank outside quotes, keys folded; a key given twice makes Postfix refuse
the whole table (no DICT_FLAG_DUP_* flag), and the check then defers too.
The writer folds keys, so two cases of one address cannot be two lines.
e2e on the stand (2026-10-14, make e2e, exit 0): the wildcard application
is refused billing@ - listed for app-list - in From: (550 5.7.1), and in
the envelope by Postfix at RCPT ("553 5.7.1 <billing@...>: Sender address
rejected: not owned by user"). go vet and go test ./... pass on the
server (23 packages).
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
|
||
|
mixandClaude Opus 5.5
|
599149bf83 |
build, deploy, inbound, dmarc, backup, e2e: cite the source of what they rely on (assumptions part C)
Comments only, no behaviour changed. Each main.cf/master.cf parameter the
build sets, the wrapper's and the reload hook's reliance on Postfix 3.7.11
(postfix-script, master.5, postconf.5, cleanup_init.c, smtpd_peer.c,
README_FILES), supervisor 4.2.5 (process.py, options.py, events.rst),
logrotate 3.21.0, OpenDKIM's config, Cyrus SASL's sasldb, rspamd 3.11.1
(milter_headers.lua, multimap, the map watcher), Docker Compose, CoreDNS
and certbot now say where that is written. Measurements already there stay
measurements ("date not recorded" where none was); hand-composed fixtures
say so.
Contradictions left for their own commits: C-P31 cleanup's size limit,
C-P42 master outside supervisor's signal, C-V5/V6 startsecs, C-V10,
C-S3/Q3 sasldb written in place, C-Q4/A1 certbot's symlinks, C-R7 rspamd
skips local clients but the signature routine does not, C-P8, C-R4, C-R12,
C-M7, C-E2, C-K8/K9, C-D1.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
|
||
|
mixandClaude Opus 5.5
|
ad3d834392 |
store, web: cite the source of what they rely on (assumptions part D)
Comments only, no behaviour changed. modernc.org/sqlite v1.60.1 (the _pragma list on every connection, busy_timeout first; extended result codes; a deferred BEGIN; multi-statement Exec) and SQLite 3.53.4's documentation (transactions, WAL, foreign keys, NOCASE, upsert, julianday, partial indexes); htmx 2.0.4 src/htmx.js; the Fetch and HTML standards for Sec-Fetch-Site, Origin, SameSite, __Host-, CSP and defer order; Go's html/template. Comments the source contradicts are corrected - the driver does not serialise writes (SetMaxOpenConns(1) does), htmx:load fires per swapped element, no-store does not keep a page out of history. Left for their own commits: D-S3 a migration two processes can both run, D-S20 recipient case, D-B3 Origin: null under Referrer-Policy: no-referrer, D-H5/H6/H7/H9 (audit M11, stage 4). Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com> |
||
|
mixandClaude Opus 5.5
|
b113e72c71 |
milter, postfix, app, domain: cite the source of what they rely on (assumptions part B)
Comments only, no behaviour changed. go-milter v0.4.1 (session.go: one session per connection, replaced after a non-continue reply; macros per stage), Postfix 3.7.11 (milter8.c, smtpd_check.c reject_auth_sender_login_mismatch(), mail_addr_find.c, showq_compat.c, mail_conf_time.c, mail_params.h), Cyrus SASL 2.1.28 (saslpasswd.c, plugins/sasldb.c: deleting a missing login is not an error), OpenDKIM 2.11.0~beta2 (opendkim-db.c, opendkim-genkey), Berkeley DB 5.3 db_dump, rspamd 3.11.0. Comments that claimed what the source denies are corrected (sasldb keeps the password as given, not hashed; an embedded newline is kept, not truncated; rspamd sorts symbols by absolute weight); fixtures say seen or composed. The sender check's From: parsing gets no citations: it is replaced by the port of OpenDKIM's dkim_mail_parse next. Contradictions left for their own commits: B-G5 cleanAddress, B-PS1 first key wins, B-PQ1 the forced-expire flag, B-PQ10/PQ11 Postfix's time units. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com> |
||
|
mixandClaude Opus 5.5
|
ce363f48dd |
logtail, health, panel: cite the source of what they rely on (assumptions part A)
Comments only, no behaviour changed. Each place the log tailer, Health and cmd/panel rely on another program now says where that is written, for the version the image runs: Postfix 3.7.11 (log_adhoc(), msg_logger_print(), smtpd_whatsup(), cleanup_milter.c, qmgr_active.c, postlogd.c, showq.c, pipe.c), logrotate 3.21.0, supervisor 4.2.5, proc(5). Measurements stay measurements and say where and when; test fixtures say whether they were seen or composed. Where the source contradicts the code the comment does not bless it but names the register entry (A-8 expiry, A-12 the ISO stamp no Postfix writes, A-15 the last recipient, A-21/A-23 sasl_username, A-33 postlogd's old inode, A-38 a socket file is not a listener, A-65 the certificate served); those are fixed in their own commits. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com> |
||
|
mixandClaude Opus 5.5
|
af130e268a |
findings of the reviews of audit stage 2b
Technical review (Opus) and security review (Fable) of 5be109b..5154bba;
reports in the session's scratchpad. Nothing above low from security; one
bug and three risks from the technical review. Migration 0009 is not
deployed, so it was extended.
- Bug: two refusals of one second swapped their NOQUEUE lines (each line
went to the newest lineless row). AttachRefusalLine now matches the
client address too (send_log.client_ip, written by the journal milter -
Postfix's line names the client and the sender, never the login: the
sasl_username part of the pattern never matched and could be fed by HELO,
security F1/F2), takes the nearest row in time and the oldest of equals,
and a line already on a row goes on no other (security F6).
- Waiting lines and recoveries lived only in memory while the saved offset
moved past them: a stop within 30 s lost them. The tracker now never saves
an offset past the oldest waiting item in its file (tracker.hold,
refusalLines.oldest) - for a recovery, the client= line that names its
login - so they are read again; the "once" guards make that harmless.
- A recovered row whose login the tailer did not see was nobody's: its
domain_id falls back to the envelope domain's.
- The fingerprint's file identity is the inode alone: the device number can
change across a host reboot and make the same log look rotated.
- A refused row carries its queue id; UpdateStatus's "newest row" now skips
refused rows, so a later message with that id does not flip it to sent.
- The alive mark: its age is read first and kept from then on, so asking
supervisord for 30 s cannot make it stale; the entrypoint removes it with
the sockets - a container restart is not the check's downtime, and a mark
from before would have counted the whole outage as "≈".
- Security F3: the sender check prints no ";" of the client's in the From:
text of its reply, and the tailer reads the envelope after the first ";"
- Postfix's own text.
- A refused row with a queue id and no stored lines says so (it was told the
mail.log text of 2.2.0); the fact's em dash is the house hyphen.
- Docs: architecture (client address, no login), guide § Health (warns on
either count), help topics log and sender-check (design
|
||
|
mixandClaude Opus 5.5
|
415334c2cc |
message page: a row the sweep closed says the panel closed it (stage 2b)
The History step of every bounced row said the receiving server refused
the message permanently or Postfix gave up, "the reason is in the delivery
log below" - for a row the reconcile sweep closed (reason "reconcile")
neither is known and the log has nothing (review of stage 1). Such a row's
step is now "Closed by the panel - Postfix no longer held the message and
mail.log never said what became of it ... rather than show a delivery it
cannot evidence." The help topic log names the case (design
|
||
|
mixandClaude Opus 5.5
|
ac985d03a8 |
sender check: connections and messages counted apart; downtime estimated when supervisord is silent (stage 2b)
Following the design commit
|
||
|
mixandClaude Opus 5.5
|
65d916f26a |
logtail: a refusal the panel did not get to write is written from mail.log (stage 2b)
test / test (push) Canceled after 0s
The sender check answers 550 5.7.1 and writes the refused row through a
buffer a moment later; a panel stopped in between (a rollout, a kill) left
an application refused and a log that did not know it. A refusal at MAIL
FROM needs nothing: the journal milter writes its row before it answers.
- logtail.refusalLines remembers each queue id's login from smtpd's
"<id>: client=..., sasl_username=<login>" (4096 ids at most) and, on
cleanup's "<id>: milter-reject: END-OF-MESSAGE ...: 5.7.1 From: <line> -
<why>; from=<...> to=<...>", waits 30 s and hands it to
store.RecoverRefusal - an INSERT ... WHERE NOT EXISTS a refused row of that
queue id within two minutes: the row the panel wrote, or one recovered
before, means nothing is written. It is read from the saved offset on, so
a kill is covered as well as a clean stop - which a flush of the buffer at
shutdown would not.
- Migration 0009 (not deployed yet) gains send_log.recovered; the row is
written with the line's time, the login's domain id, reason "from", the
From: line as the check printed it (to the last " - " before ";").
- The message page: a fact "Written: from mail.log after a restart: the
panel stopped before it wrote the refusal".
Tests: store (not written when the panel wrote it; written once however
often); logtail (the login from client=, the From: line with a " - " in it,
nothing before the wait is over); view (the fact). e2e
refusal_recovered_from_the_log: a From: refusal, the panel stopped, its row
deleted and the tailer's saved place moved before the message's first line
(python3 in the container, as the panel's user - a kill cannot be timed to
that moment from outside); started again, the row is back, recovered, once.
Also fixed: handlers TestMessagePageOfARefusedMessage still expected the
old sentence; it failed on the server after
|
||
|
mixandClaude Opus 5.5
|
e166a2ebd6 |
refused messages: the page shows what Postfix wrote (stage 2b, 0009)
The Delivery log box of every refused row said "This message never reached
the queue, so Postfix wrote no delivery lines for it" - untrue: Postfix
wrote a line for each. The owner, 2026-10-13: the box is required on every
message page. Pictured for him first (dev/shots, the refused variant of
out-message: wrong From with its queue id lines, rate-limited with the
NOQUEUE line); the variant has no mockup screen of its own - the guard ties
screens to pages, and its outline is out-message's.
- Wrong From: cleanup gives the message a queue id before the headers are
judged; the sender check reads {i} at end-of-headers and writes it on the
refused row. Its lines - stored under the id like any other's - are on the
page, but only those from 10 minutes before to 1 minute after the refusal
(store.SendLogLines for a rejected row): Postfix frees the id at once, and
a later message with it lends nothing. No mail.log fallback for a refused
row, for the same reason.
- Rate-limited and wrong client: refused at MAIL FROM, no queue id. Migration
0009 adds send_log.refusal_line; logtail.refusalLines matches smtpd's
"NOQUEUE: milter-reject: MAIL from ...; from=<...>" (not port 25's, not
CONNECT/EHLO) and store.AttachRefusalLine puts it on the newest refused row
of that sender (and login, when the line names it) within two minutes; a
line whose row is not written yet waits up to 30 s (retried at each pass).
- The page: the queue id fact says "given, the message was not queued" or
"none - refused at MAIL FROM, before one is given"; the box's caption is
"The line from mail.log for this refusal" for the latter; a row from
before says the line was not kept. LogLineLevel colours a refusal by its
reply: 5xx red, 4xx amber (IPv6 clients included). The id in the fact is
not muted as in the picture: the kit has no muted inline.
Tests: store (attach by sender, login, time; not a queued row, not one an
hour off; a refused row's lines are its own minutes); logtail (attach, wait
and retry, port 25 and other lines ignored); milter (the refused row carries
{i}); view (facts, caption, line levels); handlers (the line on the page,
the sentence without it). e2e refusals_show_postfixs_line: a client-address
refusal made there, and the pages of a wrong From, a rate-limited and a
wrong client refusal each show Postfix's line.
schema-migrations row 9, CHANGELOG [Unreleased], audit plan line ticked,
roadmap 25/62.
Evidence on the test server for this tree: make e2e exit 0 (the new case
in 5.4 s); go test ./... ok but for two schema tests, fixed after the run
(the schema at 0009) - store tests ok locally; the full run follows before
the push.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
|
||
|
mixandClaude Opus 5.5
|
c452606175 |
logtail: the deferred counter remembers a place in the log, not a second (stage 2b)
test / test (push) Canceled after 0s
The counter passed over, while catching up after a start, every warning no later than the last one counted - by time; mail.log stamps whole seconds, so a second warning of that second, later in the file, was lost. - The tailer gives each line its place (Place: the file's fingerprint and the offset just past the line): follow tracks the bytes the reader hands over, readRotated likewise in mail.log.1. Run hands tr.line to the counter. - store.AddSenderCheckDeferred(at, n, place) counts, moves the last-deferral mark (never back - an upsert WHERE) and saves the place (settings sender_check_counted_place) in one transaction. - Catching up, a warning in the file of the last one counted is passed over by place (end <= counted end); in another file - a rotated log read again - by time, as before. The fingerprint carries the inode, so the place stays true across a rotation. Tests: two warnings of one second, an unclean stop between them, all lines read again - each counted once; the store saves the place with the count and keeps the latest deferral time. Existing counter tests pass a zero place (time rule). CHANGELOG [Unreleased], audit plan line ticked, roadmap 24/62. Evidence: go vet ./... clean; go test ./internal/store, ./internal/logtail ok (but for the Windows-only test); the server run follows before the push. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com> |
||
|
mixandClaude Opus 5.5
|
4678015bbd |
logtail: a log rotated while the panel was down is read to its end (stage 2b)
At start, a saved offset whose fingerprint was not the current file's meant "start on the new file": what Postfix wrote between the panel's stop and the rotation was never read - delivery statuses, stored lines, refusals, inbound journal rows, the deferred counter. - tracker.resume marks the case (behind, prev); follow calls readRotated before the new file: it opens mail.log.1 (delaycompress leaves it plain), checks by fingerprint that it is the file the place belongs to, reads it from the saved offset to its end through handle (catching up: the once-protections apply) and flushes the stored lines. - Not found, not that file, or the offset past its end (rotated twice): logged, with the saved place's time - from when nothing was read. - The fingerprint gains the file's identity (device.inode, fileid_unix.go; empty elsewhere): two logs that begin alike are two files - the review-2026-08-followups row on collisions. A fingerprint saved before (hash alone) still matches by its hash (sameLog), so the first start after the upgrade resumes where it was. - store.LogtailState reads back updated_at. Tests: stop, append, rotate, append, start - both halves read, once each, the earlier line not again; a mail.log.1 that is another file is not read from the place; sameLog with an old and a new fingerprint. CHANGELOG [Unreleased], audit plan line ticked, roadmap 23/62. Evidence: GOOS=linux go vet ./internal/logtail clean; go test ./internal/logtail ok on Windows but for TestFollowTailsAndRotates; the server run follows before the push. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com> |
||
|
mixandClaude Opus 5.5
|
ef18ef21a0 |
findings of the reviews of audit stage 2
test / test (push) Canceled after 0s
Technical review (Opus) and security review (Fable) of a59e3d4..cb5a4a2;
reports in the session's scratchpad. Nothing critical, high or medium; no
outright bug. Fixed:
- F2, F3 (0008 backfill, before it runs on the live database): rows are
filled by the rule new rows follow - the application's domain, then the
name's - and only by a domain and application that existed when the row
was written (created_at); the inbound log likewise. Test: the migration's
UPDATEs re-run on prepared rows (before the domain, after, null sender,
application gone, a stranger's name). Dry run on a copy of the live
database: 68/68 and 10/10 rows filled, none disagreeing with its login's
domain, integrity ok.
- V2, F4: an application's statistics and level-2 count are its login's
rows of its domain (appByLogin), and StatsWindow starts no earlier than
the entity's creation; the message page's "may send as" only for a login
still of the row's domain. Test: a login taken again in another domain
starts at zero.
- F7: MAIL FROM:<> is held to the domain limit of its login
(ApplicationByLogin fills DomainName).
- V3: an imported DKIM selector is held to DNS labels (validate.Selector).
- F5, F6: an imported auto limit's numbers are dropped and recalculated
right after the import; its mode is stored trimmed.
- V6: the store refuses a role that is neither global nor domain.
- F9: SaveUser keeps the stored hash when no new password is given (the
form no longer writes back the hash it read). F10: a reach naming a
domain deleted meanwhile is ErrNoSuchDomain, answered on the form.
- V4, F12: from-lines-other.txt - lines naming another sender ({other}) in
a quoted string, a quoted pair, capitals, a comment, an encoded word, a
quoted local part; the unit test wants each refused, the e2e
(lines_naming_another_sender_never_signed_for_it) wants each refused and
none signed for it. The passed list gained a.{DOMAIN}, a comment after
the angle address, a quoted local part, and "b@ {other}" - passed by
design (security.md), and the e2e shows it signed for the own domain.
- F13: e2e login_in_other_capitals_does_not_authenticate - AUTH with the
login upper-cased gets 535.
Not changed, written down: V1/F1 - L8 is half closed (refused or passed
still tells an application which domains are hosted here); security.md's
accepted risks and the CHANGELOG say so, and the owner is asked whether to
refuse every foreign address in a display name (changes the 2.4.0 rule).
F11 - deleting a domain rewrites its rows' domain_id (seconds at 1e5 rows),
accepted at this size. F8 (test cannot tell write-first from read-first;
the reviewer verified it with two stores, 30/30), V5, V7, P2: noted.
Evidence on the test server for this tree: go vet clean, go test ./... 23
packages ok, make e2e exit 0 with the three new cases.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
|
||
|
mixandClaude Opus 5.5
|
5747f0b25c |
sender check: the reply names no other domain (L8); which key signs, tested (stage 2)
L8: a From: line naming another sender of the server was refused with
"it names another sender of this server, <domain>" - a confirmation, to
whoever tries names, that the domain is hosted here. The reason now stops
before the domain; the line as written stays in the reply (the client's own
text) and on the refused message's page (header_from). Unit test: the reason
of every refusal names none of the server's other domains; e2e: the reason
of each sender-check refusal does not contain other.e2e.test.
Which key OpenDKIM signs with (owner, 2026-10-13): the check and OpenDKIM
read From: with two parsers, and their agreement was argued, never run.
internal/milter/testdata/from-lines-passed.txt holds the lines the check
passes for an application - bare and bracketed, quoted names with a comma,
escaped quotes and an address inside, an extension, three encoded charsets,
two folded lines, comments (nested too), a stranger's address in the name -
as Go-quoted strings with {domain}. The unit test wants each passed; the new
e2e case passed_from_lines_signed_for_their_domain_only sends each as app1
and wants every DKIM signature at the sink to be for the sender domain,
with one valid. All fifteen were. "(the shop) a@domain" - a comment before
the address, valid RFC 5322 - is refused by the check ("it must be one
address"); strict, not a hole, so it is not in the file. Groups are refused
too.
Audit plan lines ticked (two) with what was found, CHANGELOG [Unreleased],
roadmap 21/62.
Evidence on the test server for this tree (with
|
||
|
mixandClaude Opus 5.5
|
a618ffb0e4 |
send log: rows belong to a domain by id (audit L12, 0008)
A domain user's log, the message page, a domain's statistics, the level-2 count and the notifier's refusal count picked send-log rows by domain name, so a domain deleted and re-added under its old name showed its earlier mail to the next assignee and went on counting it. Migration 0008 (not run on the live database yet, so extended here): send_log.domain_id and inbound_spam_log.inbound_domain_id, REFERENCES ... ON DELETE SET NULL, filled once by name from the registry as it is at migration time, each with an (id, created_at) index. - Writes: InsertQueued / InsertRejected set domain_id to the domain of the row's login (loginDomainID) - the journal milter and the sender check's refusal rows both go through them; the inbound journal sets inbound_domain_id from the inbound domain's name at insert. - Reads: SendLogFilter.DomainIDs (was Domains, by name) and InboundSpamFilter.DomainIDs are the scope; the global role still reads everything. The message page checks row.DomainID against the user's domains. DomainSendStats and CountMessages keep their name argument and select domain_id = (SELECT id FROM domains WHERE name = ?). The notifier joins send_log.domain_id to the domain and its application. - DMARC reports stay by name on purpose (a report is about the DNS name). Tests: store - scope by ids; a re-added domain's log, level-2 count and statistics hold its own row only, the global log both; inbound scope by ids. Handlers - the new assignee gets 404 for the old row's page and does not see it in the log, the global role sees both. Tests that wrote rows for a login that is no application now register it (a row is a domain's through its application). Schema at 0008 with the two columns and indexes. schema-migrations row 8, CHANGELOG [Unreleased], audit plan line ticked, roadmap 19/62. Evidence: go vet ./... clean; go test ./... ok locally but for the two Windows-only tests; the server run follows before the push. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com> |
||
|
mixandClaude Opus 5.5
|
e469034477 |
import: the export file held to the forms' rules before anything is created (audit M10)
test / test (push) Canceled after 0s
domain.Import trusted the file for everything but the domain name, the selector and the password: DMARCRua was not validate.Email'ed, client IPs not parsed, a rate limit's mode, window and count not checked - a window of 0 counted nothing, a garbage IP made the application refuse every client. handlers.checkDomainImport runs in HandleImportDomain after the domain name check and before Import, and refuses the file with "Invalid <what> in export file: <the form's own text>": the report address (validate.Email), each rate limit (parseRateLimitMode; auto - checkAutoMultiplier, 0 meaning the default as an empty field does; manual - checkManualLimit, which the form now uses too: count > 0, <= the level-1 backstop, window > 0; both 0 is an export without a limit of its own), each application's client addresses and the legacy ones on its rate limit (parseIPList, written back as the form stores them; allow-list on with none is refused). Login, sending addresses and password stay checked in app.ImportApplication. Tests: a good export passes and its addresses come out normalised; nine broken fields each refused with their message; a refused file posted to the handler creates no domain. CHANGELOG [Unreleased], audit plan line ticked, roadmap 18/62. Evidence: go vet clean; the import, rate-limit and application tests of ./internal/web/handlers ok; the server run follows before the push. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com> |
||
|
mixandClaude Opus 5.5
|
f9e75de8d1 |
store: users saved in one transaction, the last global kept inside it (audit M7, M9)
M7: DeleteUser counted the global users and then deleted; the handler's demotion and CreateGlobalUser checked and then wrote. MaxOpenConns(1) serialises statements, not pairs: two globals deleting or demoting each other at once could leave none. Now: - DeleteUser: a transaction that deletes first (taking the write lock) and then requires a global left (requireAGlobal) - ErrLastGlobal rolls back. - SaveUser: likewise for a demotion, the count inside the transaction after the write. The handler's early count stays only to answer before hashing a password; SaveUser decides, and ErrLastGlobal maps to the same message. - CreateGlobalUser: INSERT ... SELECT ... WHERE NOT EXISTS (SELECT 1 FROM users), one statement. M9: submitUserUpdate made five writes in a row (UpdateUser, sessions, SetUserRole, ClearUserReach with its error only logged, SetUserReach). store.SaveUser(UserChange) writes name, hash, e-mail, role and reach in one transaction (a global user's reach cleared); sessions end after it commits. The create path had the same flaw - user, reach and then e-mail in three writes - and now goes through AddUser, one transaction; CreateUser stays as a wrapper for its many test callers. SetUserRole is gone; writeUserReach is the reach write inside a transaction. Tests: store - two globals demoting, and deleting, each other from two goroutines: exactly one refused, one global left; a refused SaveUser (taken name, a reach to no domain, nothing assigned) changes nothing; promoted to global loses the reach; AddUser with a bad reach leaves no user; four setups at once create one user. Handlers (the users pages, direct): the whole change saved, back to domain with a reach; the last global not demoted and the typed rename not written; a domain user with nothing assigned refused, nothing changed; create with e-mail and reach. CHANGELOG [Unreleased], audit plan lines ticked (two), roadmap 17/62. Evidence: go vet ./... clean; go test ./internal/store ok; the users tests of ./internal/web/handlers ok; the server run follows before the push. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com> |
||
|
mixandClaude Opus 5.5
|
60c7b85f84 |
findings of the reviews of audit stage 1
test / test (push) Canceled after 0s
Technical review (Opus) and security review (Fable) of 9efc9c7..a59e3d4; reports in the session's scratchpad. Fixed: - B1 (verified bug): store.UpdateStatus read before writing in its transaction; a commit by another connection in between (the dmarc-ingest process opens the same file) failed it at once with "database is locked" and the delivery line was lost. The first statement is now a write. - R3: a delivery line updated every row of its queue id and recipient; it now updates only the newest (Postfix reuses short ids). The sweep closes by row id (QueuedDelivery.ID, CloseUnreported(id)). - R4: a contradicted sweep keeps a mark, reason "reconcile-contradicted", instead of losing which row it was. - R1: contradicted cannot see a lost line (never read late). New counter reconcile_missed: at close, the sweep looks for the row's delivery line in mail.log and mail.log.1 (QueueLines). 0007 edited - it has run on no database but throwaway stands. - R2: queued_over_lifetime cannot measure "what removing the sweep would leave" while the sweep runs; its text now says rows stuck despite the sweep, and the rule uses closed - missed - contradicted. - cert-reload (R5, R6, R7, security medium/low): both files copied to .new and the copies checked as a pair before both are renamed; acts on the copy only when postconf says Postfix reads it; a failed reload is retried next pass; umask 077; says so when the mounted files cannot be read. The e2e case asserts Postfix reads the copy, the copy equals the mount, the key is 600 root. - Security low: a recipient with a quoted local part - the journal cut "a b"@x at the space, deliveryRe cut "a>b"@x at the ">" - matched no delivery line and was closed by the sweep. cleanAddress ends a bracketed address at the first ">" outside quotes; deliveryRe takes the recipient up to the ">, " before relay= (test fixtures given the relay= every real delivery line has). - Security low: postqueue -p's error carries at most 512 bytes of output. - Nits: addDeliveryLogDay refuses a column not on its list; the truncation test checks its message. - To stage 2b (page wording, mockup-first): the history step of a bounced row says the receiving server refused it, false for a row the sweep closed. To the owner when stage 2b is reached: since M13 the "deferred" figure counts whole messages too, which "Connections turned away" would not name. Found while verifying: the e2e's main positive path failed once with "553 not owned by user app1" - an smtpd idle from before the panel's reload took the connection in the reload's second. Measured on the stand (one-off probe, not committed): sending 0, 10, 40, 110 s after the save (two seconds at the soonest) always met an smtpd started after the reload. The e2e now waits for the sender map after each application it creates (waitForSenderMap: AUTH, MAIL, RCPT, RSET - no message, no row), and the level-1 case uses it instead of its own loop. Docs: architecture § Queue reconcile, schema-migrations (rows 7 and 8, chain head 8), CHANGELOG, audit plan (stage 1 note, the measured window, a new stage 2b line: 62 lines), roadmap 15/62. Evidence on the test server for this tree: go vet clean, go test ./... 23 packages ok, make e2e exit 0 (renewed certificate, hung panel, level 1 at 49/60). Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com> |
||
|
mixandClaude Opus 5.5
|
f51f5db245 |
app: logins folded to lower case, unique without regard to case (audit H4, 0008)
Postfix's reject_sender_login_mismatch compares the SASL login with the sender map without regard to case - seen on the e2e stand this time (a one-off probe, not committed: map rewritten to "APP1", client authenticated as "app1", RCPT accepted). The panel stored logins as typed under a BINARY UNIQUE, so "Ops" and "ops" were two applications to the panel and one sender to Postfix: a tenant could send with another tenant's envelope sender. - app.foldLogin: Service.Create lower-cases the login (owner: fold, do not refuse); CreateWithSettings rolls back the stored (folded) login. - ImportApplication keeps the login as written: folding it would rename the SASL account and break the imported application's working credentials; the index below refuses a case clash there too. - Migration 0008 (DDL): CREATE UNIQUE INDEX idx_applications_login_nocase ON applications (login COLLATE NOCASE). The column's own BINARY UNIQUE stays (SQLite cannot change a collation without a table rebuild, which a migration transaction cannot do with foreign keys on). Checked on the live installation first: 4 logins, no case clash; "Quieter3C" keeps its case. 0008 is not run anywhere yet; send_log.domain_id joins it later in stage 2. Tests: "Ops" created as ops, SASL account and map binding in lower case, OPS in another domain and an imported oPs refused; the store refuses ops beside Ops; schema at user_version 8 with the index. CHANGELOG [Unreleased], audit plan line ticked, roadmap 15/61. Evidence on the test server for this tree (with the probe): go vet clean, go test ./... 23 packages ok, make e2e exit 0. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com> |
||
|
mixandClaude Opus 5.5
|
db36aa2c0e |
logtail: count the message a hung sender check defers in cleanup (audit M13)
Seen on the e2e stand, the panel frozen with SIGSTOP once DATA was
answered: the client gets 451 4.7.1 after 30 s, and the warning that names
the socket is cleanup's, not smtpd's:
postfix/cleanup[1339]: warning: milter unix:/run/selfpost/sender.sock:
can't read SMFIC_HEADER reply packet header: Connection timed out
postfix/cleanup[1339]: 7BCAF3C75DA0: milter-reject: END-OF-MESSAGE ...
4.7.1 Service unavailable - try again later; ...
The counter read smtpd lines only, so Health's "messages deferred" missed
the one case that tells a hung panel from a stopped one. milterLineRe now
takes smtpd and cleanup; cleanup's "milter unix:<socket>:" warning counts a
message each time (one cleanup process serves message after message, and the
first warning ends the message); smtpd keeps once per connection. The
composed fixture is replaced by the seen lines, with two messages through one
cleanup process and OpenDKIM hung in cleanup (not counted).
e2e panel_hung_mid_message_is_counted: freezes the panel after DATA's 354,
wants a 4xx at the end of DATA, logs what mail.log says, and waits for
Health's today count to grow. kill through sh -c: the image has no kill(1).
CHANGELOG [Unreleased], audit plan line ticked, roadmap 13/61.
Evidence on the test server for this tree (with the next commit's docs):
go vet clean, go test ./... 23 packages ok, make e2e exit 0 -
renewed_certificate_picked_up 1.2 s, panel_hung_mid_message_is_counted
31.7 s.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
|
||
|
mixandClaude Opus 5.5
|
63520f1678 |
logtail: deliveryRe anchored at the head of the line (audit M12)
The pattern that writes a row's status matched anywhere in a line: a port-25 reject line ending helo=<ID: to=<victim@...>, status=sent> parsed as (ID, victim, sent). parseDelivery now takes the stamp off with SplitTimestamp and matches `^host program[pid]: <id>: to=<...>, ... status=`, as queueLineRe and the inbound patterns do. Tests: the probed HELO line, the same text in a quoted sender on a qmgr line, the text after an smtpd line's queue id - none parse; a syslog-stamped line as Postfix 3.7 writes it parses. The existing cases unchanged. CHANGELOG [Unreleased], audit plan line ticked, roadmap 11/61. Evidence: go test ./internal/logtail ok but for TestFollowTailsAndRotates (Windows); the server run of the stage follows before the push. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com> |
||
|
mixandClaude Opus 5.5
|
ffefa9d6df |
delivery log: measure the sweep before deciding its fate (audit H3, 0007)
Migration 0007 (DDL only): delivery_log_days, a row per UTC day, kept a year, shown nowhere - read off a copy of the database after a month by the decision rule written in the audit plan. - reconcile_closed: the sweep adds the rows it actually closed. - reconcile_contradicted: store.UpdateStatus, in one transaction, counts the rows with reason "reconcile" a delivery line now speaks of, writes the log's status and clears the reason - so a re-read line counts once. - tailer_blind_seconds: follow() tracks stretches in which the file is gone or reopening fails (logtail.blindness); stored when a stretch ends and every minute while it lasts; under 3 s not counted (the rotation's tick between rename and create loses nothing). - queued_over_lifetime: every sweep overwrites the day's level with the rows queued past LongestLifetime + MaximalBackoff (a message Postfix still works on stays queued until its first retry after the lifetime). - retentionLoop prunes days older than a year. Tests: store (sums per day, the level overwritten, under a second nothing, prune; a contradicted sweep counted once and its reason cleared; count of queued rows); logtail (the sweep counts what it closed; countStuck records the level; blindness - a one-tick gap not counted, 75 s stored as 60 + 15 from where the last part ended). Schema tests at user_version 7. schema-migrations.md: chain head 7 (it still said 5 since 2.4.0), row 7. CHANGELOG [Unreleased], audit plan line ticked, roadmap 10/61. Evidence: go vet ./... clean; go test ./internal/store ok, ./internal/logtail ok but for TestFollowTailsAndRotates (Windows); the server run of the stage follows before the push. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com> |