Commit graph

60 commits

Author SHA1 Message Date
a9a4e912f4 fix: remove the fast timer for start_watch
All checks were successful
/ build (push) Successful in 5m6s
2026-06-04 19:41:48 +02:00
37b9676dd4 feat: shorten the time before restart in kubernetes yaml
All checks were successful
/ build (push) Successful in 5m5s
2026-06-04 19:32:44 +02:00
750e8dcbf0 feat: move start_watch into trait definition, to do that use async_traits
Some checks failed
/ build (push) Has been cancelled
2026-06-04 19:31:20 +02:00
8a7c3f5203 chore: add constant for lables used in kubernetes 2026-06-04 16:47:29 +02:00
bb60a21050 Merge pull request 'Support running insde kubernetes' (#5) from pr-kube-yamls into main
All checks were successful
/ build (push) Successful in 6m10s
Reviewed-on: #5
2026-06-03 23:29:40 +02:00
8da0f78696 feat: handle internall address differntly if running inside a cluster 2026-06-03 23:27:42 +02:00
876b568a54 Merge branch 'main' into pr-kube-yamls 2026-06-03 15:58:16 +02:00
c4395ddc10 fix: Role permission groups and service label matcher 2026-06-03 15:57:47 +02:00
2244bc6c0d wip 3 2026-06-03 15:44:42 +02:00
b5f9c5e81c fix: could not initilaize kube-rs inside cluster, should infer config now
All checks were successful
/ build (push) Successful in 5m1s
2026-06-03 13:25:14 +02:00
beba05a56d feat: add forgejo actions yaml to auto build and push a container image
All checks were successful
/ build (push) Successful in 4m59s
2026-06-03 11:50:47 +02:00
a8311bb8ca wip #2 2026-06-03 11:37:17 +02:00
80b264331e wip 2026-05-29 17:40:33 +02:00
5dc61a0f12 feat: make OpaqueError FanctTrace more legible 2026-05-29 13:10:05 +02:00
0e2afd0772 feat: polish the OpaqueError get_span_trace function looks 2026-05-29 11:07:59 +02:00
1bce912a1a feat: polish logging if the server is unavailable 2026-05-29 10:42:24 +02:00
1229127101 feat: add better logging span to process_connection 2026-05-29 10:40:51 +02:00
ee7a05152e feat: cleaner logging
- less clutter with unusable `span` info (on my personal preference)
- less errors logged when being scanned (ex.: after status response
    don't log errors with ping packets, scanners usually close
    the connection)
- implemented more detailed OpaqueError fields
    - `tracing::Level`
    - error `kind`
2026-05-28 18:00:29 +02:00
efe080a9e7 feat: add localized tmux server 2026-05-28 17:59:28 +02:00
a5da8a4e15 doc: update README.md env variables to a better format 2026-03-07 13:58:38 +01:00
8a9ae9f735 feat: add Text component format or raw JSON text disconnect messages 2026-02-03 21:04:21 +01:00
6fab222d1e feat: add dependency derivation to flake output, so it can be pinned
If you write nix build .#deps ,then the result link gets added
as a gc root so it doesn't get garbage collected.
2026-02-03 21:03:20 +01:00
e403d93dc2 fix: login_username_extract span added for extended logging information 2026-01-29 10:24:51 +01:00
bebca902df feat: call start_watcher when connecting to an already running instance
(used to be only when starting the server)
2026-01-29 10:24:51 +01:00
c70527ded5 fix: fix all instances of not_found span instrumenting, and change the error to trace as an error 2026-01-19 16:34:17 +01:00
d531ba59a0 feat: add login_start packet parsing for username extraction 2026-01-19 16:34:08 +01:00
c697449b20 feat: add basic docker image build 2026-01-17 11:19:24 +01:00
737ff501ab chore: update flake.lock 2026-01-17 11:17:24 +01:00
0d8c04aea3 fix: #1 ; next_state could be 0, which is invalid 2026-01-16 09:59:08 +01:00
a58024c192 fix: server_watcher don't use unwrap and use the external port quering the server 2026-01-13 13:32:26 +01:00
e67afc23df feat: don't rebuild dependencies(of the crate) if the hash changes
It used to rebuild all of the dependencies of the crate, because I
embed the hash in the program, and when commiting there is a new hash.
("revision" or "hash", it is just the git hash)

Now it only rebuilds the final crate, so there is caching again. :)
2026-01-12 13:57:05 +01:00
39b1a84c55 fix: PacketLengthInvalid did not correctly return at any time
this was due to the value being coerced to usize and then
panicing in debug with integer subtract underflow
2025-12-17 18:44:20 +01:00
822330ef87 feat: add connection filters; and config struct;
also move sanitize_addr to mc_server module
2025-12-14 12:56:21 +01:00
3dcf2f03a8 fix: server_watcher pool JoinHandles and check them 2025-12-12 02:54:28 +01:00
7f6bd226a7 fix: if server couldn't be found and the next packets are bad, only emit one line 2025-12-12 03:37:37 +01:00
1454945f32 feat: better looking motd, and kubernetes native motd handling 2025-12-11 23:16:05 +01:00
402e315a93 feat: use port value of handshake to determine where to route the data 2025-12-11 21:54:19 +01:00
6c56b9e0e3 fix: modded(neoforge 1.20.1) address string handling 2025-12-11 16:41:13 +01:00
2eb9adc61a feat: add some botched old address support 2025-12-11 02:26:08 +01:00
1d9d205d8c feat: gracefully handle status request, when server doesn't exist 2025-12-11 01:43:51 +01:00
0d5fec173d feat: gracefully handle servers, which could not be found with the addr 2025-12-11 01:10:58 +01:00
5bfff0a081 feat: add watcher which checks if the joined servers became empty and stops them 2025-12-10 23:10:37 +01:00
138429948a fix: reworked, but not renamed 2025-12-10 19:27:39 +01:00
2892cc3e7b feat: add abstraction for proxying logic + small fixes
and i dont wanna break this up into multiple commits.
2025-12-10 19:25:59 +01:00
5233570675 fix: don't abort parsing packets if they have larger than 0 size 2025-12-07 17:14:58 +01:00
f1c1b2c122 feat: improve packet::parse tracing 2025-12-07 12:16:28 +01:00
7461b8ea30 fix: replace todo!() with Ok(()) 2025-12-06 20:04:14 +01:00
594e91b8a1 chore: startup logging polish; BIND_PORT env var setting added 2025-12-04 00:25:07 +01:00
4cf3d5aea0 feat: add an Error type to Packet::parse
instead of error logging manually inside `parse`,
return a Result with a custom enum
2025-12-03 22:58:21 +01:00
92dfbd490c chore: small cleanup 2025-12-03 22:58:01 +01:00