TriliumNext Server install

[doobes]

Hi guys,

I realized that my existing Trilium server (on a RaspBerryPi) wasn't using my mapped NAS volume to store data in.

So seeing that the TriliumNext version is up and running, I've decided to try that.

I've got a shared volume on my NAS at //192.168.10.10/trilium/. I can mount to it outside of docker easy peasy. I've tried this with and without a mount outside of Docker. Doesn't seem to make any difference.

I'm using a docker compose file and it looks like this:

services:
    trilium:
        image: 'triliumnext/notes:latest'
        container_name: trilium
        ports:
            - '8080:8080'
        volumes:
            - /NAS-trilium_data:/home/node/trilium-data
        restart: unless-stopped
volumes:
  NAS-trilium_data:
    driver_opts:
      type: "cifs"
      o: "username=username,password=password,uid=1000,gid=1000,file_mode=0777,dir_mode=0777,noexec,nosuid,nosetuids,nodev,vers=3.0"
      device: "//192.168.10.10/trilium/"

Running docker compose up, it all works without error.

But it's not using the NAS volume for data storage.

The compose stuff for the mount comes from here:

https://random.blackpacket.net/2021/01/10/docker-nfs-and-cifs-volumes.html
Any thoughts?

thanks

chris

[doobes]

Ok,

This is kinda wierd

It put the data files in the root of the drive:

It it's obviously not mounting the drive

[eliandoran]

@doobes ,

Can you try removing the forward slash of the volume mount?

        volumes:
-             - /NAS-trilium_data:/home/node/trilium-data
+             - NAS-trilium_data:/home/node/trilium-data

See https://forums.docker.com/t/docker-compose-mount-samba-volume/132407/6

[doobes]

Ok.

One step forward, one step sideways as they say.

Did as you suggested. It didn't help that my password was incorrect in docker-compose.yml.

Sort all that out and now Trilium is erroring out:

✔ Network pi_default            Created                                                                                                                                          0.2s
✔ Volume "pi_NAS-trilium_data"  Created                                                                                                                                          0.0s
✔ Container trilium             Created                                                                                                                                          2.2s
Attaching to trilium
trilium  | No USER_UID specified, leaving 1000
trilium  | No USER_GID specified, leaving 1000
trilium  | (node:9) ExperimentalWarning: Importing JSON modules is an experimental feature and might change at any time
trilium  | (Use `node --trace-warnings ...` to show where the warning was created)
trilium  | Generated session secret
trilium  |
trilium  |
trilium  | #
trilium  | # Fatal error in , line 0
trilium  | # Check failed: module->status() == kLinked || module->status() == kEvaluated.
trilium  | #
trilium  | #
trilium  | #
trilium  | #FailureMessage Object: 0x7fc003e7f0
trilium  | ----- Native stack trace -----
trilium  |
trilium  |  1: 0xd0e488  [node]
trilium  |  2: 0x1fb19e0 V8_Fatal(char const*, ...) [node]
trilium  |  3: 0x1384fe4 v8::internal::SourceTextModule::Evaluate(v8::internal::Isolate*, v8::internal::Handle<v8::internal::SourceTextModule>) [node]
trilium  |  4: 0xed67d4 v8::Module::Evaluate(v8::Local<v8::Context>) [node]
trilium  |  5: 0xc517f0 node::loader::ModuleWrap::Evaluate(v8::FunctionCallbackInfo<v8::Value> const&) [node]
trilium  |  6: 0xf1e378 v8::internal::FunctionCallbackArguments::Call(v8::internal::CallHandlerInfo) [node]
trilium  |  7: 0xf1eb38  [node]
trilium  |  8: 0xf1ef50 v8::internal::Builtin_HandleApiCall(int, unsigned long*, v8::internal::Isolate*) [node]
trilium  |  9: 0x189c964  [node]
trilium  | Trace/breakpoint trap
trilium exited with code 0

It's writing files to the shared volume as I can see that from the Synology DSM app.

Thoughts?

[doobes]

Same result. Trilium errors out.

docker-compose.yml file:

services:
    trilium:
        image: 'triliumnext/notes:v0.90.6-beta'
        container_name: trilium
        ports:
            - '8080:8080'
        volumes:
            - NAS-trilium_data:/home/node/trilium-data
        environment:
            - USER_UID=1000
            - USER_GID=1000
        restart: unless-stopped
volumes:
  NAS-trilium_data:
    driver: local
    driver_opts:
      type: "cifs"
      o: "username=username,password=password,uid=1000,gid=1000,vers=3.0"
      device: "//192.168.10.10/trilium"

Files are written to the mapped drive. But Trilium errors out and then repeatedly restarts.

[perfectra1n]

Yeah that certainly does seem to be the same issue as previous versions. Are you able to share the output of neofetch or some additional information about the Pi that you're using such as Model/OS/Architecture?

[eliandoran]

@doobes , for your information, the issue you have initially described seems to be #408 , which got solved by switching from SMB to NFS.

[doobes]

@doobes , for your information, the issue you have initially described seems to be #408 , which got solved by switching from SMB to NFS.

or not

1. Mount the Docker volume:

sudo docker volume create --driver local \
  -o type=nfs \
  -o o=addr="192.168.10.10,rw,nfsvers=4" \
  -o device=:/volume1/trilium \
  nfs-trilium

2. Create/run the image:
   sudo docker run -t -i -p 127.0.0.1:8080:8080 -v nfs-trilium:/home/node/trilium-data triliumnext/notes:v0.90.6-beta

docker: Error response from daemon: error while mounting volume '/var/lib/docker/volumes/nfs-trilium/_data': failed to mount local volume: mount :/volume1/trilium:/var/lib/docker/volumes/nfs-trilium/_data, data: addr=192.168.10.10,nfsvers=4: protocol not supported.

I've got my NFS permissions set to accept connections from 172.18.0.0/24.

Trouble is, I don't know exactly what ip address Docker is creating.

This is the issue with NFS mounts./

Can we please fix the CIFS issue?

[perfectra1n]

The error protocol not supported. means that you will need to drop your nfsvers=4 to nfsvers=3.

Can we please fix the CIFS issue?

PRs are more than welcome @doobes :)

[doobes]

This is my current docker-compose.yml

services:
    trilium:
        image: 'triliumnext/notes:latest'
        container_name: trilium
        ports:
            - '8080:8080'
        volumes:
            - NAS-trilium_data:/home/node/trilium-data
        restart: unless-stopped
volumes:
  NAS-trilium_data:
    driver: local
    driver_opts:
      type: "cifs"
      o: "username=usename,password=password,uid=1000,gid=1000,vers=3.0"
      device: "//192.168.10.10/trilium"

[doobes]

Ok,

Just for fun, I deleted all the Docker stuff, volumes, containers, etc.

Mounted the shared folder on the NAS as /mnt/trilium

docker-compose.yml looks like this:

services:
    trilium:
        image: 'triliumnext/notes:latest'
        container_name: trilium
        ports:
            - '8080:8080'
        volumes:
            - /mnt/trilium:/home/node/trilium-data
        environment:
            - USER_UID=1000
            - USER_GID=1000
        restart: unless-stopped

It writes the files to /mnt/trilium, but I get the same error.

[doobes]

I've some other stuff to do right now.

So I reset it to write locally so I can access the server in the interim.

Let me know what you think.

cheers,

chris

[perfectra1n]

It's working on my Raspberry Pi 4B:

[perf3ct on homelab-monitor1] ~
»  sudo docker run -p 8080:8080 triliumnext/notes:latest
No USER_UID specified, leaving 1000
No USER_GID specified, leaving 1000
(node:9) ExperimentalWarning: Importing JSON modules is an experimental feature and might change at any time
(Use `node --trace-warnings ...` to show where the warning was created)
Generated session secret
DB not initialized, please visit setup page - http://[your-server-host]:8080 to see instructions on how to initialize Trilium.
DB size: 4 KB
{
  "appVersion": "0.90.6-beta",
  "dbVersion": 228,
  "nodeVersion": "v20.15.1",
  "syncVersion": 32,
  "buildDate": "2024-09-07T18:36:34Z",
  "buildRevision": "7c0d6930fa8f20d269dcfbcbc8f636a25f6bb9a7",
  "dataDirectory": "/home/node/trilium-data",
  "clipperProtocolVersion": "1.0",
  "utcDateTime": "2024-09-07T20:44:20.047Z"
}
CPU model: Cortex-A72, logical cores: 4, freq: 1500 Mhz
Trusted reverse proxy: false
App HTTP server starting up at port 8080
Listening on port 8080
Slow 200 GET /api/health-check with 15 bytes took 14ms
Creating database schema ...
Becca (note cache) load took 1ms
Creating root note ...
Created option "revisionSnapshotTimeInterval" with default value "600"
Created option "protectedSessionTimeout" with default value "600"
Created option "zoomFactor" with default value "1.0"
Created option "overrideThemeFonts" with default value "false"
Created option "mainFontFamily" with default value "theme"
....

Here's what my neofetch shows:

[perf3ct on homelab-monitor1] ~ took 3m3s
 >  neofetch
       _,met$$$$$gg.          perf3ct@homelab-monitor1
    ,g$$$$$$$$$$$$$$$P.       ------------------------
  ,g$$P"     """Y$$.".        OS: Debian GNU/Linux 11 (bullseye) aarch64
 ,$$P'              `$$$.     Host: Raspberry Pi 4 Model B Rev 1.1
',$$P       ,ggs.     `$$b:   Kernel: 6.1.21-v8+
`d$$'     ,$P"'   .    $$$    Uptime: 21 mins
 $$P      d$'     ,    $$P    Packages: 1488 (dpkg)
 $$:      $$.   -    ,d$$'    Shell: fish 3.6.1
 $$;      Y$b._   _,d$P'      Terminal: /dev/pts/3
 Y$$.    `.`"Y$$$$P"'         CPU: BCM2835 (4) @ 1.500GHz
 `$$b      "-.__              Memory: 523MiB / 1849MiB
  `Y$$
   `Y$$.
     `$$b.
       `Y$$b.
          `"Y$b._
              `"""

[doobes]

Here is my neofetch

pi@trilium:~ $ neofetch
       _,met$$$$$gg.          pi@trilium
    ,g$$$$$$$$$$$$$$$P.       ----------
  ,g$$P"     """Y$$.".        OS: Debian GNU/Linux 12 (bookworm) aarch64
 ,$$P'              `$$$.     Host: Raspberry Pi 4 Model B Rev 1.1
',$$P       ,ggs.     `$$b:   Kernel: 6.6.20+rpt-rpi-v8
`d$$'     ,$P"'   .    $$$    Uptime: 1 day, 38 mins
 $$P      d$'     ,    $$P    Packages: 713 (dpkg), 4 (snap)
 $$:      $$.   -    ,d$$'    Shell: bash 5.2.15
 $$;      Y$b._   _,d$P'      Terminal: /dev/pts/1
 Y$$.    `.`"Y$$$$P"'         CPU: (4) @ 1.500GHz
 `$$b      "-.__              Memory: 154MiB / 907MiB
  `Y$$
   `Y$$.
     `$$b.
       `Y$$b.
          `"Y$b._
              `"""

As I indicated, it works fine writing locally.

In a previous life, I went through the effort to stand up a server for Joplin via Docker on the same Rpi that I'm using for Trilium. Similar, but different.

What I found was it would not set up correctly if there was anything in the mounted share. It behaved EXACTLY like what Trilium is doing.

I've gone through the Trilium setup several times. I've deleted all the Trilium containers, volumes, etc, etc, etc.

Yet when I go through the process again, it finds the data that was there last time.

So, obviously the server daemon is storing a copy somewhere.

I've used my best Linux-fu (which isn't a lot if I'm honest), but I can't find any files anywhere.

So my question is,

Where is the server pulling this data from? I have to figure with all my futzing about there are numerous copies and this can't be a good thing.

Thanks for all your help.

[doobes]

Okay,

So I started this all from scratch.

I've followed this guide line by line:

https://triliumnext.github.io/Docs/Wiki/apache-proxy-setup.html

I've got a domain I'm going to call trilium.yourdomain.com, that I've got certs for in

/etc/letsencrypt/live/trilium.yourdomain.com/fullchain.pem
/etc/letsencrypt/live/trilium.yourdomain.com/privkey.pem

When I get to:

sudo systemctl restart apache2, I get

Job for apache2.service failed because the control process exited with error code.
See "systemctl status apache2.service" and "journalctl -xeu apache2.service" for details.

typing in: systemctl status apache2.service

I get :

× apache2.service - The Apache HTTP Server
     Loaded: loaded (/lib/systemd/system/apache2.service; enabled; preset: enabled)
     Active: failed (Result: exit-code) since Mon 2024-09-09 16:30:31 EDT; 14s ago
       Docs: https://httpd.apache.org/docs/2.4/
    Process: 1505 ExecStart=/usr/sbin/apachectl start (code=exited, status=1/FAILURE)
        CPU: 76ms

Sep 09 16:30:31 trilium systemd[1]: Starting apache2.service - The Apache HTTP Server...
Sep 09 16:30:31 trilium apachectl[1508]: AH00526: Syntax error on line 3 of /etc/apache2/sites-enabled/trilium.yourdomain.com.conf:
Sep 09 16:30:31 trilium apachectl[1508]: Invalid command 'RewriteEngine', perhaps misspelled or defined by a module not included in the server configuration
Sep 09 16:30:31 trilium apachectl[1505]: Action 'start' failed.
Sep 09 16:30:31 trilium apachectl[1505]: The Apache error log may have more information.
Sep 09 16:30:31 trilium systemd[1]: apache2.service: Control process exited, code=exited, status=1/FAILURE
Sep 09 16:30:31 trilium systemd[1]: apache2.service: Failed with result 'exit-code'.
Sep 09 16:30:31 trilium systemd[1]: Failed to start apache2.service - The Apache HTTP Server.

I've run into this before and was never able to make this work.

This is the /etc/apache2/sites-enabled/trilium.yourdomain.com.conf:

 <VirtualHost *:80>
     ServerName http://**trilium.yourdomain.com**
     RewriteEngine on
         RewriteRule ^ https://%{SERVER_NAME}%{REQUEST_URI} [END,QSA,R=permanent]
 </VirtualHost>
 <VirtualHost *:443>
     ServerName https://trilium.yourdomain.com
     RewriteEngine On
     RewriteCond %{HTTP:Connection} Upgrade [NC]
     RewriteCond %{HTTP:Upgrade} websocket [NC]
     RewriteRule /(.*) ws://localhost:8080/$1 [P,L]
     AllowEncodedSlashes NoDecode
     ProxyPass / http://localhost:8080/ nocanon
     ProxyPassReverse / http://localhost:8080/
     SSLCertificateFile /etc/letsencrypt/live/trilium.yourdomain.com/fullchain.pem
     SSLCertificateKeyFile /etc/letsencrypt/live/trilium.yourdomain.com/privkey.pem
     Include /etc/letsencrypt/options-ssl-apache.conf
 </VirtualHost>

Thoughts?

[eliandoran]

@doobes , it appears you do not have the mod_rewrite module enabled.

Please refer to https://stackoverflow.com/questions/10144634/htaccess-invalid-command-rewriteengine-perhaps-misspelled-or-defined-by-a-m .

[doobes]

You are correct.

Perhaps this should be included in the documentation.

Thank you

[doobes]

Just as a summary,

All is not well in Mudville.

I cannot get Trilium to connect to either an NFS or CIFS mount, whether done outside of docker or inside.

The CIFS mount is broken and the NFS mount has issues with permissions that I'm not able to rectify.

Can y'all please fix the CIFS mount issues?

Thank you

[eliandoran]

Can y'all please fix the CIFS mount issues?

The CIFS mount issues are not issues with TriliumNext itself, but with the JavaScript runtime environment (Node.js) itself. We are investigating the issue in #415, maybe we can find a workaround.

[doobes]

Just for laughs,

I spun up a brand new RPi instance on a new Pi.

This was the process, and results:

pi@trilium1:~ $ sudo docker volume create --driver local
-o type=nfs
-o o=addr="192.168.10.10,rw,nfsvers=3"
-o device=:/volume1/trilium
nfs-trilium
nfs-trilium
pi@trilium1:~ $ sudo docker run -t -i -p 127.0.0.1:8080:8080 -v nfs-trilium:/home/node/trilium-data triliumnext/notes:v0.90.6-beta
Unable to find image 'triliumnext/notes:v0.90.6-beta' locally
v0.90.6-beta: Pulling from triliumnext/notes
bd2a3a2ed82d: Pull complete
36d8d8caa1e1: Pull complete
1bc680a5b1ea: Pull complete
434de69c0dbc: Pull complete
7865ae0aa84d: Pull complete
461d242cc9c9: Pull complete
c3d4b40146e3: Pull complete
20fb63033d50: Pull complete
f450fc5c076f: Pull complete
543b29e93df9: Pull complete
817e60f6a710: Pull complete
6828a0a90fc1: Pull complete
3819ce2501bf: Pull complete
4f4fb700ef54: Pull complete
Digest: sha256:9fcd81d7573fcefb630a049317f1847977e5c6d1f5e92c2b1ca64676df388b9b
Status: Downloaded newer image for triliumnext/notes:v0.90.6-beta
No USER_UID specified, leaving 1000
No USER_GID specified, leaving 1000
(node:9) ExperimentalWarning: Importing JSON modules is an experimental feature and might change at any time
(Use node --trace-warnings ... to show where the warning was created)
node:fs:1372
const result = binding.mkdir(
^

Error: EACCES: permission denied, mkdir '/home/node/trilium-data/log'
at Object.mkdirSync (node:fs:1372:26)
at file:///usr/src/app/src/services/log.js:6:8
at ModuleJob.run (node:internal/modules/esm/module_job:222:25)
at async ModuleLoader.import (node:internal/modules/esm/loader:316:24)
at async asyncRunEntryPointWithESMLoader (node:internal/modules/run_main:123:5) {
errno: -13,
code: 'EACCES',
syscall: 'mkdir',
path: '/home/node/trilium-data/log'
}

Whoever is saying a NFS mount works has some magic dust going.

[eliandoran]

@doobes , you would need to fix your permission issues in order to have it working. Maybe https://kb.synology.com/en-br/DSM/help/DSM/AdminCenter/file_share_privilege_nfs?version=7 could help you.

[perfectra1n]

Up until the container no longer complains EACCES: permission denied you're having permissions issues. NFS maps the same user ID and group ID on both sides (client/server) so you'll want to make sure that the same UID on both sides of the connection has permissions to write to the share. The container by default runs as 1000:1000.

[perfectra1n]

@doobes try the SMB mount again, but specify the mount options noperm and nobrl. That's what made it work for me :)

My full command:

mount -t cifs -o credentials=/root/nascreds.txt,nobrl,iocharset=utf8,file_mode=0777,dir_mode=0777,noperm,uid=1000,gid=1000 //192.168.2.155/testdataset /home/perf3ct/testmount/

[doobes]

Oi!

That did it!

It didn't like the nolock option, but everything else was fine.

Greatly appreciated!

chris

[perfectra1n]

That was my fault, it used to be called nolock, but now it's just nobrl (I've updated my command above). Glad I could help!

[eliandoran]

Great job, @perfectra1n .