Commit graph

8 commits

Author SHA1 Message Date
Deluan Quintão
ce484083bf
fix(server): exit with an error code when the server fails to start (#6236)
When a startup step failed (for example, the port was already in use), runNavidrome only logged the error and returned. In service mode, service.Run() kept waiting for a stop signal, so the process stayed up serving nothing and the service manager never restarted it. A plain run exited with code 0.

runNavidrome now returns the error, unless its context was cancelled by a normal shutdown. Both the plain run and the service goroutine exit with code 1 on that error. The systemd unit no longer lists 1, 2 and 8 in SuccessExitStatus, so Restart=on-failure restarts the service on exit code 1.

Fixes #6235
2026-09-27 14:18:24 -04:00
Deluan Quintão
b293b96256
refactor(persistence): stateless repositories with per-call context (#6149)
* refactor(persistence): adopt generic deluan/rest repository API

Pin deluan/rest to the refactor branch. REST-facing repository methods
take a context and return typed values. Drop DataStore.Resource and
ResourceRepository; the native API names typed repositories directly
through a per-request adapter that later commits remove.

* refactor(persistence): base repository helpers take a context

* refactor(persistence): LibraryRepository takes a context per call

* refactor(persistence): PropertyRepository takes a context per call

* refactor(persistence): UserPropsRepository takes a context per call

* refactor(persistence): TranscodingRepository takes a context per call

* refactor(persistence): ShareRepository takes a context per call

* refactor(persistence): PlayerRepository takes a context per call

* refactor(persistence): RadioRepository takes a context per call

* refactor(persistence): PlayQueueRepository takes a context per call

* refactor(persistence): Tag and Genre repositories take a context per call

* refactor(persistence): PluginRepository takes a context per call

* refactor(persistence): Scrobble repositories take a context per call

* refactor(persistence): FolderRepository takes a context per call

* refactor(persistence): Artwork repositories take a context per call

* refactor(persistence): UserRepository takes a context per call

* refactor(persistence): ArtistRepository takes a context per call

ReadAll no longer rewrites the shared sort mappings for the role filter;
it works on a per-call copy.

* test(persistence): assert artist role sort sanitization in ReadAll

* refactor(persistence): AlbumRepository takes a context per call

* test(persistence): pass the test context to album repository helpers

* refactor(persistence): MediaFileRepository takes a context per call

* refactor(persistence): Playlist repositories take a context per call

* refactor(persistence): build all repositories once per store

* refactor(core): REST repository wrappers are built once

* refactor(persistence): repositories are stateless

Remove the context field from the base repository and the per-request
REST adapter. Enable the containedctx linter so no repository can hold a
request context again.

* chore(lint): skip containedctx in test files

* refactor: share simplifications from the stateless repositories sweep

Add deleteOwnedAll on sqlRepository and use it in player/share Delete
to remove the duplicated bulk-delete loop; have Share.Repository()
return model.ShareRepository so subsonic sharing.go drops its repeated
type assertions.

* chore(core): assert REST wrappers implement Persistable

* chore: reformat imports

* perf(persistence): build repositories on first use

Each transaction store used to construct all 21 repositories up front,
paying for filter and sort mapping setup the block never touched. Fields
are now sync.OnceValue thunks, so a store only builds what it uses.

* fix(persistence): clean plugin references per deleted user

A bulk user delete that fails on a later id had already removed the
earlier rows but skipped their plugin cleanup. Cleanup now runs right
after each successful delete.

* fix(core): unload disabled plugins even when a user delete fails

A bulk delete can fail on a later id after earlier users were removed
and their plugins auto-disabled. The wrapper returned before unloading,
leaving those plugins running until the next successful delete or a
restart.

* chore(deps): pin deluan/rest to v1.0.1

Replaces the pseudo-version of the refactor branch with the tagged
release. REST error messages now name the bare type (Artist, not
model.Artist).

* test: use the spec context instead of context.Background()

Replace the context.Background()/context.TODO() calls this branch added
to tests with the spec's ctx, GinkgoT().Context(), or t/b.Context(), so
repository calls are bound to the running spec's lifetime.

* test: declare the spec context once per Describe

Set ctx from GinkgoT().Context() first in each top-level BeforeEach and reuse it, building user contexts on top of it instead of repeating inline calls.
2026-09-25 18:06:10 -04:00
Deluan Quintão
42d4363f61
fix(service): rewrite systemd service template for kardianos/service v1.3.0 (#5743)
kardianos/service v1.3.0 (shipped in Navidrome 0.63.0) replaced Go's
text/template with a small custom engine that uses bare keys ({{Description}})
instead of dotted fields ({{.Description}}). Our custom SystemdScript was
still written in text/template syntax, so 'navidrome service install' failed
with 'FATA service: unknown template key ".Description"', breaking the .deb
postinst and leaving an empty/masked systemd unit.

Rewrite the template in the new engine's syntax and add a test that validates
every key and pipeline function used in the template against the set the
library provides to systemd templates, so future engine/key drift is caught
at test time. Verified end-to-end on Linux: the previous binary reproduces
the reported fatal error and the fixed one generates a complete unit file.

Fixes #5742
2026-07-08 16:40:07 -04:00
Deluan Quintão
8f0b4930ff
refactor(conf): replace eager dir creation with lazy Dir type (#5495)
* feat(conf): add Dir type with lazy directory creation

Introduces the Dir type that wraps a directory path string and defers
os.MkdirAll until the first call to Path() or MustPath(), using sync.Once
to ensure the creation happens exactly once. Implements fmt.Stringer,
encoding.TextMarshaler, and encoding.TextUnmarshaler for config integration.
Includes Ginkgo/Gomega tests covering all methods and error paths.

* refactor(conf): replace eager dir creation with lazy Dir type

Change DataFolder, CacheFolder, Plugins.Folder, and Backup.Path from
string to Dir. Remove all os.MkdirAll calls from Load() so directories
are created lazily on first Path()/MustPath() call. Artwork folder
creation was already handled at point-of-use in image_upload.go.

Add SnapshotConfig() to conf package for safe test config save/restore
that avoids copying sync.Once inside Dir fields. Fix copy-lock vet
warning in nativeapi/config.go by marshalling pointer instead of value.

* refactor(conf): migrate tests and db init to lazy Dir type

Update all test files to use conf.NewDir() for Dir field assignments.
Ensure DataFolder is created lazily when the database is first opened
in db.Db(). Remove eager directory creation from conf.Load() tests.

* fix(conf): address review findings for Dir type

- Use os.ModePerm for DataFolder/CacheFolder (was 0700, should match
  original behavior). Add NewDirWithPerm for PluginsFolder (0700).
- Use Path() instead of MustPath() in db.Prune() to avoid logFatal
  from background cron job.
- Panic on marshal/unmarshal errors in SnapshotConfig (test helper).
- Clean up redundant String()/MustPath() calls in plugin manager.
- Remove dead code in dir_test.go.

Signed-off-by: Deluan <deluan@navidrome.org>

* fix(conf): add GoString to Dir for clean config dump output

Implement fmt.GoStringer on Dir so pretty.Sprintf shows the path
string instead of internal struct fields (sync.Once, perm, err).
Also add TODO comment to configtest about removing the indirection.

* fix(dir): improve error logging in MustPath method

Signed-off-by: Deluan <deluan@navidrome.org>

* refactor(tests): remove redundant tests for unwritable DataFolder and CacheFolder

Signed-off-by: Deluan <deluan@navidrome.org>

* fix(conf): address PR review feedback

- Ensure Plugins.Folder always uses 0700, even when user-configured
  (previously only the derived default got restrictive permissions).
- Create LogFile parent directory before opening, so LogFile paths
  inside a not-yet-created DataFolder work correctly.

---------

Signed-off-by: Deluan <deluan@navidrome.org>
2026-05-13 17:44:22 -03:00
Kendall Garner
8a19fa9991
fix(server): require additional variable to enable systemd logging (#5222)
* fix(logging): require additional variable to enable systemd logging

* use a better name
2026-03-23 18:09:59 -04:00
Kendall Garner
154e13f7c9
build: add packages for deb and rpm to release (#3202)
* support packing deb/rpm/archlinux

* .-.

* initial test

* fix postinstall, remove execstop

* bash -> sh, create toml manually if it doesn't exist (thanks debian)

* don't forget that newline

* postrm

* comments, contrib -> packaging/linux

* contrib > packaging in .goreleaser

* actually add toml

* openrc/sysv templates

* add apk. nothing else yet

* wait, we have a ntive uninstall

* fix: merge errors, move packaging to release

* chore: remove old goreleaser conf

* ci: remove `release` dependency on `docker push`

* ci: fix release version

* ci: upload packages

* ci: try to fix json file list

* ci: replace the json file list with a txt artifact

* postremove -> preremove, skip install/remove error

* actually do preremove

* better preremove

* ci: fix

* ci: fix?

* ci: clean-up

* ci: try to change labels and filenames

* ci: fix?

* ci: fix?

* ci: add `make package` target

* ci: make labels more readable

hope it doesn't break the pipeline again

* build: remove alpine and archlinux packages, for now.

---------

Co-authored-by: Deluan <deluan@navidrome.org>
2024-10-26 13:31:45 -04:00
Rob Emery
9c3b456165
feat(build): MSI installer improvements (#3376)
* feat(build): add a make target to build a msi installer locally

* Testing wrapping the executable in cmd

* build(ci): build msis in parallel

* feat(server): add LogFile config option

* Revert "Testing wrapping the executable in cmd"

This reverts commit be29592254.

* Adding --log-file for service executable

* feat(ini): wip

* feat(ini): parse nested ini section

* fix(conf): fix fatal error messages

* Now navidrome supports INI, we can use the built-in msi ini system
and not require the VBScript to convert it into toml

* File needs to be called .ini to be parsed as an INI and correct filename
needs to be passed to the service

* fix(msi): build msi locally

* fix(msi): pipeline

* fix(msi): pipeline

* fix(msi): pipeline

* fix(msi): pipeline

* fix(msi): pipeline

* fix(msi): Makefile

* fix(msi): more clean up

* fix(log): convert LF to CRLF on Windows

* fix(msi): config filename should be case-insensitive

* fix(msi): make it a little more idiomatic

* Including the latest windows release of ffmpeg into the msi
as built by https://www.gyan.dev/ffmpeg/builds/ (linked
to on the official ffmpeg source)

* This should version independent

* Need bash expansion for the * to work

* This will run twice, once for x86 and once for x64, I'll make it cache
the executable for now as it'll be quicker

* Silencing wget

* Add ffmpeg path to the config so Navidrome knows where to find it

* refactor: download ffmpeg from our repository

* When going back from the "Are you ready to install?" it should go back to the
Settings dialogue that you just came from

* fix: comments

---------

Co-authored-by: Deluan <deluan@navidrome.org>
2024-10-22 19:32:56 -04:00
Rob Emery
768160b05e
feat: Windows MSI installer and service support (#3125)
* First version/rough layout of the required wix to build an MSI that embeds everything

* Don't need revision number

* produced exe from existing build process is navidrome not Navidrome

* Adding Kardianos wrapper around Cobra so the callbacks are handled
automatically (this is basically only for windows)

* Adding pointless check to shut up lint for now

* make format

* Revert disabling npm tidy

* Using Kardianos always will result in the application hanging so it
needs only be wrapped to handle the callbacks if it's being used in
the service context, otherwise use cobra directly

* Copying in service installation etc from https://github.com/navidrome/navidrome/pull/2295

* Under Linux this installs a user service (I don't think this is
correct, but lets get this working first). User units/services
cannot depends on system units, so previously this bombed out
with Exit Code 5.

* Under Windows we can install both the x86 and x64 builds, they
will install to different folders, but previously they would
overwrite the service as they were both called Navidrome. Now,
it will install 2 services. This will still be weird/broken as
they will attempt to listen on the same port, however uninstalling
the "wrong" arch will not cause the "right" one to be partially
uninstalled anymore

* Reverting changes to the context as they don't really seem necessary anyway

* Need to consistently name the service

* Fixing broken context

* The included files should be removed when the app is uninstalled

* Reverting back to the original context here, I don't think
it makes any difference to running under kardianos

* Let's see what we have immediately available

* OK, the build takes ages so let's just try and do the whole thing in one go, maybe we'll get lucky

* Need -r on directory copy, plus we'll probably need to install wixl

* No sudo cmd, so I assume this runs as root

* WORKSPACE!

* Moving the version to be a single variable, we'll probably be able to pull it from the github tag or whatever

* Might as well put the msi in the right folder, it's tidier

* Writing the version number into the msi, from the output of goreleaser

* Using jq to parse the goreleaser metadata, so need to install it

* MSI only supports numerical version numbers, so I'll make the "snapshot" version .1 minor patch greater

* -r or --raw (on newer versions) means we don't get the "" around the value

* Running as a user service I think makes limited sense for this

* Will now ask for configuration settings during install.

MSI/WiX only supports writing out INI files, Toml is almost
INI compatable, except that the INI needs to write out a section
first, so we need to have a script to strip that off.

We are forced to display a License.rtf file by the UI so I think
the build process should probably rename the default licence file
and that will suffice.

Uninstalling works cleanly, howvever upgrades seem to leave the old
version installed in "programs and features" currently.

Adding the UI has introduced a requirement for WiX 0.103

* Updating the build to include --ext ui for the new config ui

* Configuration dialog should not display for upgrades as the config file
is already written

* Making description consistent with the systemd service and making
the build process produce the required License.rtf

* Fixing " non-constant format string in call to fmt.Errorf (govet)"

* Its a string, not an int; read better.

* Wixl 103 is required for --ext ui, so we need 24.04

* OK this is still installing Wix 0.101, maybe it all needs to be 24.04?

* Switching the builds back to ubuntu-latest (22.04 at current)
as it runs on a custom container, it's actually debian anyway
Moving msi build into its own job so it can run on 24.04 so
we have access to wixl 0.103 for --ext ui support

* Forcing build

* Whitespace fix

* Adding sudo I guess

* Gotta checkout as well

* Adding debugging for when there's soemthing wrong with the paths

* Adding more ls to see if the output has worked

* The msi's are in subdirs

* Actually they're in the ./wix directory

* Still can't find these msi's?

* I think that was being treated literally previously

* No idea why this isn't working, give it a relative path instead?

* Making explicit on the dialogue that the configuration file will be
where the installation dir is

* The lint keeps failing and it's just getting in the way so I'll
turn this off for now and we'll edit out this commit from the merge

* Cutting more out of the build to get more stuff out of the way

* Need to increase the width to fit the text in

* Calling everything License.rtf, presumably one of them is correct

* I am pretty sure the License.rtf loading is broken under Wixl; so
let's just bypass the EULA from the UI which is a nicer experience
for the users anyway

* This needs to be after WelcomeDlg now the Eula isn't displayed

* You're supposed to be able to use <WixVariable> to override the
location that the bmp's are loaded from, I can't get this to work
under wixl so I'm guessing given that the ui extension is new, it
hasn't been implemented with that in mind. So we'll hack it by
overwriting the files installed with the package.

* We should make this less brittle so when wixl is updated it still works

* Re-enabling the lint and tests etc

* Improving the scaling quality and removing borders from images to
tidy them up a tad

* Pretty sure this isn't necessary as MY_PROPERTY will always be false

* Without publishing this event, we can't continue to the next dialogue
however I think we should be able to get away without the property

* Refactoring out the duplication so we only have one service
defined and we can run that either way

* Pushing the Interactive check into the root commmand? Feels like it
is probably getting closer to the right place at least

* go tidy

* OK this didn't work under windows, I'm guessing it's because
it's lacking all the metadata about the service it needs to
report back to Windows on.

* We need to run service execute now so that the windows
service will behave (hopefully)!

* Lint

* go tidy

* Renaming service to "navidrome" rather than "Navidrome" as this
is the filename that systemd writes and it's unusual to have
capital letters in service names under Linux.
Switching to use service execute for Linux to mirror Windows

* Need to provide the arguments to append

* Without passing the context around, the DB isn't closed gracefully
so we end up with with .db-shm and .db-wal files for recovery

* We should log fatal rather than outputting directly to stdout

* go tidy

* refactor: small nitpicks

* fix: terminate service gracefully

---------

Co-authored-by: Deluan Quintão <deluan@navidrome.org>
2024-10-01 19:40:53 -04:00