jcoffey-dev 7c80a12d75
ci / fork-checks (pull_request) Successful in 18s
ci / build (pull_request) Successful in 17m2s
Task manager: release task locks on stop, recheck claims held elsewhere
A cluster rehearsal (PostgreSQL + NATS) left index tasks pending well
past the one-hour task lock after the node that claimed them was stopped
or killed. The exact cause there isn't confirmed; this closes every path
found in the task manager that stretches a takeover past the lock, or
keeps a task claimed without running it:

- A graceful stop never released the locks it held, so every task the
  node had claimed stayed blocked for an hour. The server now tracks the
  locks it holds (common::ipc::TaskLocks) and, once the shutdown signal
  arrives, stops claiming and releases them before exiting.
- A node that failed to claim a task (another node held it) set its own
  local hold for a full lock lifetime from that scan. If the holder
  claimed it just after the scan began, or ran on a clock ahead, that
  hold ran out a moment before the lock did and was set for another
  hour: two hours in all. Such claims are now tried again every five
  minutes (a twelfth of the lock lifetime), and the task manager wakes
  up for them: before, a node without a coordinator could sleep up to
  five minutes past the recheck, or until something else woke it.
- A worker that panicked took its task type down on that node for good,
  while the scan kept claiming that type's tasks and failing to hand them
  over, re-taking each lock as it expired and so starving every other
  node of them. Each batch now runs on a task of its own; a panic is
  logged, the batch's locks are released and the worker carries on. A
  failed hand-over releases the lock too.
- A claimed task the worker couldn't read, or found gone, kept its lock
  for the hour. It is released.
- An IndexDocument task for a file (not indexed) returned no result,
  which shifted every later result in the batch onto the wrong task in
  update_tasks. It returns Ignored. Nothing queues such a task today.

The lock lifetime stays one hour; it now lives per server so the tests
can shorten it.

store::task_locks::task_lock_tests plays a second node by writing its
locks straight into the in-memory store: tasks it claimed and abandoned
run here once its locks expire, including locks that outlive this node's
view of them, and a graceful stop hands this node's locks back at once
and claims nothing more. It passes on RocksDB, SQLite and PostgreSQL.
With the old recheck it fails.
2026-09-24 08:19:05 -07:00
2026-09-18 10:21:56 -07:00

inbuxa

A complete mail and collaboration server, every feature included, under the AGPL


inbuxa is a mail and collaboration server: JMAP, IMAP, POP3, SMTP, CalDAV, CardDAV and WebDAV, in one Rust binary, with ihasmail as its web front end. It is a fork of Stalwart. Project site: inbuxa.org. Documentation: docs.inbuxa.org.

Stalwart ships some features only in a paid Enterprise Edition: multi-tenancy, masked email, undelete and others. inbuxa ships everything to everybody under the AGPL-3.0, rebuilding those features independently and without using any of Stalwart's Enterprise code.

What's different from Stalwart

  • Every feature, one edition. No license key, no edition checks, no upsell. See docs/spec/SPEC.md §4 for the features being rebuilt, and docs/spec/features/ for each one's specification.
  • Webmail and administration by ihasmail, as a separate service that can run beside the server or elsewhere. Stalwart's own web interface is removed, so there's no web front end on the mail host.
  • Clean-room rebuilds. Enterprise-only code is stripped from every upstream release before it's imported. The rebuilt features are written from specifications that use only public sources (docs/spec/SPEC.md §3).

How the fork is kept

Upstream releases arrive as stripped snapshots, never with upstream's git history, which contains Enterprise code. tools/fork/strip.py builds each snapshot on top of upstream's own ossify.py, then verifies it independently. The report for every import is in docs/fork/strip-reports/. See docs/spec/SPEC.md §2.

Building

cargo build --release -p inbuxa          # the binary is target/release/inbuxa
docker build -t inbuxa .                 # or the container image

Settings are read from INBUXA_* environment variables. An existing Stalwart install's STALWART_* variables aren't read: the server stops at startup and names each one to rename. New installs keep their data in /var/lib/inbuxa and logs in /var/log/inbuxa. Existing installs keep the paths their configuration already names, so none of their data moves.

License and credits

inbuxa is free software under the GNU Affero General Public License, version 3.

It is a fork of Stalwart, copyright © Stalwart Labs LLC, modified by Coffey Labs in 2026. Upstream's copyright notices are kept on every file they cover, and every upstream file this fork changed says so in its header, under the notice it came with. Stalwart's files are dual-licensed AGPL-3.0-only or Stalwart's Enterprise License, and inbuxa takes them under the AGPL-3.0 only. A few of those files also carry code from other projects under MIT or BSD licenses, which stays under those licenses; THIRD-PARTY.md lists it with its notices. "Stalwart" is Stalwart Labs' name. inbuxa isn't affiliated with or endorsed by Stalwart Labs.

The inbuxa mark reuses ihasmail's cat-and-envelope artwork.

S
Description
Imported from github.com during the 2026-09-20 standup (local dir: inbuxa-server)
Readme
15 MiB
2026-09-23 06:11:03 +00:00
Languages
Rust 96.6%
Python 2.1%
HTML 1.1%