prosody-hg: plugins/mod_smacks.lua history

mod_c2s, mod_smacks: Refactor session/connection cleanup

Tue, 21 Apr 2026 16:22:43 +0100

mod_c2s, mod_smacks: Refactor session/connection cleanup

mod_smacks was preventing mod_c2s cleanup by directly modifying the mod_c2s
sessions table. It did this to avoid mod_c2s trying to destroy the session
when the connection was closed.

The code that does IP counting in mod_limits was experiencing counter drift,
likely because of mod_smacks disassociating the session from the connection
before the c2s-closed event is fired.

This change has mod_c2s skip the session cleanup if it sees the 'hibernating'
property is set on the session. This is a less invasive way of performing the
same thing. Importantly, c2s-closed still gets to run and fire with the
original session object.

It might be desirable to remove the 'hibernating' check here and switch to a
more generic event-driven design (where mod_smacks can inhibit cleanup by
adding an event handler). This change was deemed too large for a minor
release, but may be made in trunk.

mod_smacks: Remove extra optional from sm element

Tue, 02 Sep 2025 00:50:25 +0200

mod_smacks: Remove extra optional from sm element

Since XEP-0198 version 1.6.3, the optional and required children of sm in
stream:features have been removed entirely. They already weren’t specified
before, and were only leftover from a very old version in the XML Schema.

Fixes #1963

mod_smacks: Run session destruction code in session's async context

Thu, 21 Nov 2024 17:02:55 +0000

mod_smacks: Run session destruction code in session's async context

This is the Right Thing and should fix various potential issues and race
conditions.

mod_smacks: Fix previous commit for case of no session found

Tue, 19 Nov 2024 00:57:24 +0100

mod_smacks: Fix previous commit for case of no session found

mod_smacks: Prevent traceback on attempting to resume destroyed session

Tue, 19 Nov 2024 00:53:01 +0100

mod_smacks: Prevent traceback on attempting to resume destroyed session

This shouldn't be possible, but seems to happen anyway after net.server
somehow calling the close handler twice, which also should not happen.

Merge 0.12->trunk

Sat, 02 Nov 2024 22:33:38 +0100

Merge 0.12->trunk

mod_smacks: Destroy timed out session in async context (fixes #1884)

Sat, 02 Nov 2024 18:10:25 +0100

mod_smacks: Destroy timed out session in async context (fixes #1884)

Prevents ASYNC-01 due to storage interactions in a timer.

Also considered modifying mod_c2s to allow passing arbitrary closures
into its runner thread but this seems like a big step away from the
current code for just this module.

Also considered creating a dedicated runner in mod_smacks, but ensuring
continuity across module reloads might be tricky.

We could further improve this in the next major version.

mod_smacks: Adjust buckets for resumption age statistic

Fri, 16 Feb 2024 00:15:36 +0100

mod_smacks: Adjust buckets for resumption age statistic

Given that there are recommendations floating around recommending 24
hours session lifetime, having buckets up to 10 minutes wouldn't be
useful in that case.

Would be nice if we had some way to automatically assign suitable number
series for buckets, scaled to what the configuration might be.

plugins: Use integer config API with interval specification where sensible

Mon, 17 Jul 2023 01:38:54 +0200

plugins: Use integer config API with interval specification where sensible

Many of these fall into a few categories:
- util.cache size, must be >= 1
- byte or item counts that logically can't be negative
- port numbers that should be in 1..0xffff

plugins: Switch to :get_option_period() for time range options

Sun, 16 Jul 2023 20:49:33 +0200

plugins: Switch to :get_option_period() for time range options

Improves readability ("1 day" vs 86400) and centralizes validation.

Merge 0.12->trunk

Tue, 04 Apr 2023 08:55:15 +0200

Merge 0.12->trunk

mod_smacks: Replace existing watchdog when starting hibernation

Sun, 02 Apr 2023 10:40:48 +0200

mod_smacks: Replace existing watchdog when starting hibernation

There shouldn't be one here but if there is, for some reason, it's
better to close it than have it around to wake up and possibly try to
destroy the session.

mod_smacks: Fix stray watchdog closing sessions

Sun, 02 Apr 2023 10:37:44 +0200

mod_smacks: Fix stray watchdog closing sessions

Unsure exactly how this happens, but sometimes a watchdog appears to
close a session that isn't hibernating, or hasn't hibernating long
enough.

plugins: Prefix module imports with prosody namespace

Fri, 24 Mar 2023 13:15:28 +0100

plugins: Prefix module imports with prosody namespace

Merge 0.12->trunk

Thu, 19 Jan 2023 21:14:31 +0100

Merge 0.12->trunk

mod_smacks: Log something when hibernation starts

Thu, 19 Jan 2023 20:59:28 +0100

mod_smacks: Log something when hibernation starts

Will hopefully save future confusion about sessions being destroyed when
they are in fact not.

mod_smacks: Factor out resumption token table key generation

Sun, 13 Nov 2022 19:44:53 +0100

mod_smacks: Factor out resumption token table key generation

So that happens in a single place, where it can be changed easier.

Merge 0.12->trunk

Mon, 12 Dec 2022 07:10:54 +0100

Merge 0.12->trunk

mod_smacks: Only track resumable sessions

Fri, 18 Nov 2022 16:18:47 +0100

mod_smacks: Only track resumable sessions

Required due to track_session() having moved here

mod_smacks: Disable resumption behavior on s2s

Sun, 13 Nov 2022 18:16:59 +0100

mod_smacks: Disable resumption behavior on s2s

Since resumption is not supported on s2s currently, there is no point in
allocating resumption tokens. The code that removes entries from
session_registry is only invoked for c2s sessions, thus enabling
resumable smacks on s2s adds an entry that never goes away.

mod_smacks: Change boolean attribute from '1' to 'true' for compatibility

Fri, 07 Oct 2022 11:35:56 +0100

mod_smacks: Change boolean attribute from '1' to 'true' for compatibility

Conversations 2.10.10 and earlier expect this to be literally 'true' and don't
recognise '1'. This leads to it not attempting resumption with Prosody at all
since this change was introduced in 36ba170c4fd0.

Thanks to Zash for noticing, debugging and diagnosing this issue.

This issue is fixed in Conversations commit 052c58f3 (unreleased at the time
of writing).

Backed out changeset 1bc2220cd6ec

Mon, 03 Oct 2022 12:55:11 +0200

Backed out changeset 1bc2220cd6ec

The use of the error helpers creates an `` child element
containing the error condition. This is however not allowed as per
XEP-0198, which specifies that the error condition is to be a direct
child of the `` stream management element.

This has triggered a fun reconnect loop in aioxmpp where it was
reported by a user [1].

[1]: https://github.com/horazont/aioxmpp/issues/382

mod_smacks: Set session flag during successful enable

Sat, 03 Sep 2022 21:20:29 +0100

mod_smacks: Set session flag during successful enable

mod_smacks: Add type field to results so actions can be later distinguished

Sat, 03 Sep 2022 21:20:07 +0100

mod_smacks: Add type field to results so actions can be later distinguished

mod_smacks: Set session.smacks after sending to fix traceback

Mon, 29 Aug 2022 15:58:51 +0100

mod_smacks: Set session.smacks after sending to fix traceback

...with opportunistic writes enabled.

mod_smacks: Use new :add_error() in last remaining error result construction

Mon, 29 Aug 2022 15:48:07 +0100

mod_smacks: Use new :add_error() in last remaining error result construction

mod_smacks: Split enable handling to stages, to allow easier SASL2 integration

Mon, 29 Aug 2022 15:45:52 +0100

mod_smacks: Split enable handling to stages, to allow easier SASL2 integration

mod_smacks: Split resumption into multiple stages, to simplify ISR integration

Fri, 26 Aug 2022 19:07:36 +0100

mod_smacks: Split resumption into multiple stages, to simplify ISR integration

This will allow us to return the success/failed as part of the SASL2 response,
and *then* perform the stanza sync as a second step.

mod_smacks: Don't close resuming session when failed due to overflow

Fri, 26 Aug 2022 17:28:06 +0100

mod_smacks: Don't close resuming session when failed due to overflow

mod_smacks: Long overdue cleanup of resumption code, fixes some old TODOs

Fri, 26 Aug 2022 17:04:15 +0100

mod_smacks: Long overdue cleanup of resumption code, fixes some old TODOs

Merge 0.12->trunk

Mon, 13 Jun 2022 19:17:36 +0200

Merge 0.12->trunk

mod_smacks: Fix #1761 by setting a flag earlier

Sun, 12 Jun 2022 01:50:33 +0200

mod_smacks: Fix #1761 by setting a flag earlier

This ensures that the flag is set even if the pre-drain callback is
called from send(), as would be the case if opportunistic writes are
enabled.

mod_smacks: Indicate that bounces are generated by the server

Fri, 27 May 2022 14:54:32 +0200

mod_smacks: Indicate that bounces are generated by the server

Could arguably be implied by 'recipient-unavailable' since if it was
available, this error wouldn't happen.

Merge 0.12->trunk

Fri, 27 May 2022 15:08:55 +0200

Merge 0.12->trunk

mod_smacks: Bounce unhandled stanzas from local origin (fix #1759)

Fri, 27 May 2022 14:45:35 +0200

mod_smacks: Bounce unhandled stanzas from local origin (fix #1759)

Sending stanzas with a remote session as origin when the stanzas have a
local JID in the from attribute trips validation in core.stanza_router,
leading to warnings:
> Received a stanza claiming to be from remote.example, over a stream authed for localhost.example

Using module:send() uses the local host as origin, which is fine here.

mod_smacks: Fix bounce of stanzas directed to full JID on unclean disconnect

Fri, 27 May 2022 12:05:47 +0200

mod_smacks: Fix bounce of stanzas directed to full JID on unclean disconnect

Fixes #1758

Introduced in 1ea01660c79a

In e62025f949f9 to and from was inverted since it changed from acting on
a reply to acting on the original stanza (or a clone thereof)

Unsure of the purpose of this check, you don't usually send stanzas to
your own full JID. Perhaps guarding against routing loops?

The check was present in the original commit of mod_smacks,
prosody-modules rev 9a7671720dec

mod_smacks: Remove debug log references to timer (not used anymore)

Thu, 26 May 2022 19:24:01 +0200

mod_smacks: Remove debug log references to timer (not used anymore)

Cuts down on noise as well

Merge 0.12->trunk

Thu, 26 May 2022 17:48:59 +0200

Merge 0.12->trunk

mod_smacks: Fix to use current method of counting acked stanzas

Thu, 26 May 2022 17:38:55 +0200

mod_smacks: Fix to use current method of counting acked stanzas

Fixes #1757

These places seem to have been left since e62025f949f9

The logic around expected_h in should_ack() misbehaved, always comparing
with 0 + unacked instead of acked + unacked.

mod_smacks: Use session logging for remaining log messages

Sat, 21 May 2022 13:11:25 +0200

mod_smacks: Use session logging for remaining log messages

For consistency and easier correlation of session events.

mod_smacks: Factor out some convenience functions

Fri, 11 Feb 2022 16:09:42 +0100

mod_smacks: Factor out some convenience functions

Those lines are long and the risk of mistakes if another one needs to be
added seems high, but lower when factored out like this.

mod_smacks: Initialize queue before sending

Sun, 15 May 2022 16:12:34 +0200

mod_smacks: Initialize queue before sending

Setting the .smacks field enables code paths that expects the queue to
be present. The queue is initialized in wrap_session_out(). With
opportunistic writes enabled this happens immediately on .sends2s(), so
the sending must happen before OR after these two lines, not in
the middle.

mod_smacks: Improve activation of smacks on outgoing s2s

Sun, 24 Apr 2022 16:17:32 +0200

mod_smacks: Improve activation of smacks on outgoing s2s

Using a timer was a hack to get around that stream features are not
available at the right time and sendq stanzas were stored as strings
so could not be counted properly. The later has now been fixed and the
former is fixed by recording the relevant stream feature on the session
so that the correct version of XEP-0198 can be activated once the
connection has been authenticated and is ready to start.

mod_smacks: Tweak resumption age buckets towards multiples of 60

Fri, 04 Feb 2022 22:11:14 +0100

mod_smacks: Tweak resumption age buckets towards multiples of 60

This seems like the thing to do for time, which is usually divided into
divisors divisible by 60, or multiplied by multiples of 60

mod_smacks: Log error to silence luacheck

Fri, 31 Dec 2021 00:59:10 +0100

mod_smacks: Log error to silence luacheck

mod_smacks: Limit "old" session resumption counters

Fri, 31 Dec 2021 00:48:38 +0100

mod_smacks: Limit "old" session resumption counters

Doing this when creating a whole new session seems reasonable because it
is already expensive and this is when it may be adding to the old
session store, while a successful resumption should be plus-minus zero.

mod_smacks: Record timestamp with persisted counters

Fri, 31 Dec 2021 00:18:13 +0100

mod_smacks: Record timestamp with persisted counters

For future cleanup routine when people inevitably complain about this
data being stored there forever

mod_smacks: sprinkle some metrics on it

Mon, 27 Dec 2021 16:16:53 +0100

mod_smacks: sprinkle some metrics on it

mod_smacks: Compact code using new stanza API

Wed, 22 Dec 2021 22:13:03 +0100

mod_smacks: Compact code using new stanza API

mod_smacks: Use more compact resumption tokens

Wed, 22 Dec 2021 18:47:45 +0100

mod_smacks: Use more compact resumption tokens

UUID seems like insane overkill for something user-scoped and not
security-sensitive. All that is needed is to avoid conflicts among what
should be relatively long-lived sessions.

mod_smacks: Add future todo

Mon, 20 Dec 2021 00:21:04 +0100

mod_smacks: Add future todo

The vision: All the queueing and counting and session replacement logic
in core, with mod_smacks only hooking this up to the XEP-0198 protocol.

mod_smacks: Only include max resumption time when resumption is activated

Mon, 20 Dec 2021 00:18:47 +0100

mod_smacks: Only include max resumption time when resumption is activated

mod_smacks: Fix typos

Mon, 20 Dec 2021 00:06:47 +0100

mod_smacks: Fix typos

mod_smacks: Fix error on shutdown if loaded on Component

Mon, 20 Dec 2021 00:04:32 +0100

mod_smacks: Fix error on shutdown if loaded on Component

mod_smacks: Fix typo (thanks codespell)

Sun, 19 Dec 2021 15:00:22 +0100

mod_smacks: Fix typo (thanks codespell)

mod_smacks: Avoid log noise when a non-smacks session is closed

Sun, 19 Dec 2021 14:59:27 +0100

mod_smacks: Avoid log noise when a non-smacks session is closed

mod_smacks: Check if session was really sleeping before logging message

Sat, 18 Dec 2021 15:45:06 +0100

mod_smacks: Check if session was really sleeping before logging message

It could have been resumed without going into hibernation first, i.e.
when the client notices the disconnect before the server, or if it
switches networks etc.

mod_smacks: Fix traceback when bouncing unacked stanzas

Sat, 18 Dec 2021 13:58:33 +0100

mod_smacks: Fix traceback when bouncing unacked stanzas

Errors sent from handle_unacked_stanzas() should usually not be sent to
the session itself, but if one is, it should not be queued.

mod_smacks: Fix duplicate sleep behavior when replacing a session

Sat, 18 Dec 2021 13:48:57 +0100

mod_smacks: Fix duplicate sleep behavior when replacing a session

Fixes that an extra watchdog was set, leaking the previous one, which
went on to do behave as if the session times out.

mod_smacks: Remove references to connection on hibernation

Sat, 18 Dec 2021 12:50:53 +0100

mod_smacks: Remove references to connection on hibernation

Still having the connection on the session may cause unintentional
behavior, such as the session being treated as if connected, even tho
the connection has been closed.