Hi, I've been running my at-home self-hosted Cloudron for (a large number of) years quite happily. It's got 5-6 regular users. Basic configuration:

• Intel Nuc i7 2.70GHz, 64GB RAM
• 250GB m2 nvme drive mounted as / (OS, "critical" app Data Directories as defined by my users)
• 1 x 500GB internal SATA ("uncritical" apps Data Directories)
• 1 x 6TB USB disk (dedicated to backups)
• 1 x 1TB USB disk (dedicated to users Nextcloud storage as mounted storage volume)
• 1 x 4TB USB disk (multimedia mounted to various apps as mounted storage volume)

I've been slowly running out of storage on / over the years, to the point that I've got 20-30GB free.

I'm constantly tweaking but I'm loosing the battle with platformdata - 21.59 GB and docker 20.48 GB over the years. If anyone's got suggested silver bullets, I'm all ears!

So I need to work out how to migrate from my 250GB / to something bigger.

I'd appreciate advice on how to achieve this.

So far I'm gotten this:

1. Move all app Data Directories to one of my large USB disks
2. Do a complete backup to....my USB backup drive?
3. Power down
4. Swap out my nvme 250GB / drive and replace with larger (I'm thinking 1TB?) nvme
5. Install same <?> Ubuntu OS and Cloudron version?
6. Do a move to another server?

I'd appreciate advice, corrections, on this please. Downtime isn't an issue per-say as it's for a small company but my users are impatient!

Thanks.

@nebulon Thanks thats clear.

@nebulon said in Is it possible to change where files are hosted to another volume?:

Surfer only has one data root, which is set to /app/data/public within the app on Cloudron. To serve up files from other volumes/disks you can just change the data directory of the whole app https://docs.cloudron.io/apps/#data-directory

Hi @nebulon and that then moves the app config files to the external mount (in my case the USB disk). Right?

Happy New Year all!

I've got a question about moving where Surfer app files are stored.

Context
(This Cloudron is a home hosted instance. App data directory is on the same disk as Cloudron OS)

I want to host some large files (e.g. +4GB operating system images) with the Surfer app. The files are too large to host in the Data Directory (Default location /app/data/public on the same physical disk as the machine OS).

I've added an external Mounts (a large USB disk I use for files) - /media/10tbfilestorage. The surfer app has RW access.

I can't see a way or a config file to change the storage directory to the 10TB storage. I've checked the app documentation but can't see a way.

What happens now
When I upload files to the Surfer app, they're uploaded to the default location /app/data/public.

What I want to happen
When I upload files to the Surfer app, they're uploaded to the USB disk mount.

Is it possible to move that file storage location? Or do I need to move the whole Data Directory from the /appdata/public location to the /media/10tbfilestorage/ location?

Thanks.

I'm not sure where the issue is here:

I'm trying to update the books I have with a cover image. There seems to be a few ways:

1. Search for cover image

Search possible images from Internet, screenshot:

2. Select image, screenshot:
   

I get an error when I try to apply image:

Failed to download image from url: EACCES: permission denied, open '/media/entertainment/Music_6TB/Audiobooks & Radio Programmes/Surrounded By Idiots (Thomas Erikson)/cover'

The permissions for this folder right now is www-data:

Logfile output:

Aug 05 18:28:31 [2024-08-05 17:28:31.974] ERROR: [CoverManager] Download image file failed for "https://m.media-amazon.com/images/I/711-u61tyVL.jpg" EACCES: permission denied, open '/media/entertainment/Music_6TB/Audiobooks & Radio Programmes/Surrounded By Idiots (Thomas Erikson)/cover' (CoverManager.js:129)

2. Upload book cover image

1. Choose local image to upload, screenshot:

2. Upload image, screenshot:
   

3. Fails with error

I'm really not sure what this error means <title>Cloudron App Error</title>, screenshot:

Log output:

Aug 05 18:31:02 [2024-08-05 17:31:02.029] FATAL: [Server] Unhandled rejection: ReferenceError: path is not defined, promise: Promise { <30>1 2024-08-05T18:31:02+01:00 apps 97366adc-c5d0-4186-99ac-f83c26a98bbf 2880069 97366adc-c5d0-4186-99ac-f83c26a98bbf - <rejected> ReferenceError: path is not defined <30>1 2024-08-05T18:31:02+01:00 apps 97366adc-c5d0-4186-99ac-f83c26a98bbf 2880069 97366adc-c5d0-4186-99ac-f83c26a98bbf - at /app/code/server/managers/CoverManager.js:97:64
Aug 05 18:31:02 at runMicrotasks (<anonymous>) <30>1 2024-08-05T18:31:02+01:00 apps 97366adc-c5d0-4186-99ac-f83c26a98bbf 2880069 97366adc-c5d0-4186-99ac-f83c26a98bbf - at processTicksAndRejections (node:internal/process/task_queues:96:5) <30>1 2024-08-05T18:31:02+01:00 apps 97366adc-c5d0-4186-99ac-f83c26a98bbf 2880069 97366adc-c5d0-4186-99ac-f83c26a98bbf - at async CoverManager.uploadCover (/app/code/server/managers/CoverManager.js:96:21) <30>1 2024-08-05T18:31:02+01:00 apps 97366adc-c5d0-4186-99ac-f83c26a98bbf 2880069 97366adc-c5d0-4186-99ac-f83c26a98bbf - at async ApiRouter.uploadCover (/app/code/server/controllers/LibraryItemController.js:205:16) <30>1 2024-08-05T18:31:02+01:00 apps 97366adc-c5d0-4186-99ac-f83c26a98bbf 2880069 97366adc-c5d0-4186-99ac-f83c26a98bbf - } (Server.js:171)
Aug 05 18:31:02 => Ensure permissions

3. Upload image the audiobook file directory

1. Upload image file to directory, screenshot:
   

2. Apply image, screenshot:
   

3. Fails with error, screenshot:
   

Again, I don't see why `Failed to download image fro, IRL: EACCES: permission denied, open..." error would be caused.

All the book directories, and all files have permissions for www-data.

Anyone seen this before? As shown, logfiles permission is denied, but permissions are I think correct. Thanks.

Great, thanks @nebulon. I was pretty sure the answer was yes, but I wanted to confirm.

Edit: since I can delete them backup dirs, I assume I can also delete the associated snapshot:

me@CLOUDRON:/var/backups# ls -la snapshot/
total 15263272
drwxr-xr-x 2 yellowtent yellowtent        4096 Jan  2  2022 .
drwxrwxrwx 8 root       root              4096 Jul 17 06:25 ..
-rw-r--r-- 2 yellowtent yellowtent 15629218346 Jan  2  2022 app_3fe158b6-7c19-4d38-880c-5811175b868f.tar.gz

I've searched the forum but haven't yet found an answer.

It's funny, the only recurring work I have with my Cloudron instance relates to backups, and that's only because of diskspace. Cloudron has given me hours a month back and paid for itself many times over!

So to my issue....

I've a Cloudron instance (CLOUDRON) doing backups to a directory on another server (BACKUP), via SSHFS. It works very well usually.

Checking earlier, I've had successful backups for at least a month.

I've been looking through CLOUDRON disk usage and found old backups (~70GB). See below:

total 2580
drwxrwxrwx  8 root       root         4096 Jul 17 06:25 .
drwxr-xr-x 13 root       root         4096 May  7  2019 ..
drwxr-xr-x  2 yellowtent yellowtent   4096 Dec 24  2021 2021-12-24-230000-685
drwxr-xr-x  2 yellowtent yellowtent   4096 Dec 25  2021 2021-12-25-010000-633
drwxr-xr-x  2 yellowtent yellowtent   4096 Jan  1  2022 2022-01-01-191853-674
drwxr-xr-x  2 yellowtent yellowtent   4096 Jan  1  2022 2022-01-01-200000-628
drwxr-xr-x  2 yellowtent yellowtent   4096 Jan  2  2022 2022-01-02-200000-686

Since my CLOUDRON VPS diskspace is limited, and I have successful backups (with snapshots) on my BACKUP server, I'd like to remove these backups listed above.

Can I safely delete these 2-3 year old backups from /var/backups, and the associated snapshots? If not, is SCPing them (and the snapshots) over to the BACKUP server a safe alternative?

Thanks.

@murgero Thanks for the reply. I did think that might be possible but I wasn't sure. In which case, I might try it myself over the next few days and see how it goes. I might be back to the forum with questions if I get stuck.

Hea @timconsidine great to hear your licensed too! I'd be very interested to hear what you think about running it on the LAMP stack.

I'm currently trying to listen to KE5EE in Florida on 20M. QSB noise is 7+!

What modes do you operate? Any interest in DMR? 73s de EI8FDB/M5FDB

In terms of common Cloudron usage this is probably niche, but there are 3M+ amateur radio (aka ham radio) operators in the world, me included!

For those unfamiliar - amateur radio operators need to log contacts they make with other amateur radio operators. They do this in their logbook. This used to be a physical paper logbook, like this one.

Nowadays it's done in software, usually installed locally, but there are some "cloud" based services.

Cloudlog is one of those hosted services. I've looked through their repo and they've got a Docker instructions.

Would anyone be able to say how difficult it would be to package Cloudlog?

I'd be happy to do some of the legwork (opening issues for necessary questions in their repo, etc) to see what's involved.

Description

Cloudlog is an open-source PHP & MySQL based amateur radio logging application, allowing you to log contacts via a web browser on any device and platform.

This is an ideal general-purpose logging application, supporting HF to Microwave, it can even interface with your radio via CAT, sync logs from WSJT-X & if you are a satellite operator integrates with SatPC32.

Website

https://www.magicbug.co.uk/cloudlog/

Repo

https://github.com/magicbug/Cloudlog

@froodle said in New UI Sneak peak and review:

I dislike our grouping but I think that one doesnt make full sense either... I am not sure what the perfect order of things is at the moment.

Hi @froodle, your question here is about information architecture - how to structure and group the information (functions) in your menu.

In order to get a better idea of what groups to put those functions into, you could do a card sort exercise with your users.

Card sorting exercises help software makes by asking the users for how they expect things to be organised. It takes the stress of you, the software maker.

https://www.nngroup.com/articles/card-sorting-definition/

This won't give you 100% the right answer, but it will help you organise it better than you could do one your own.

I hope that helps.

If you'd like to talk about card sorting, let me know and I'd be happy to help you.

Happy New Year all.

I had a disk failure over Christmas on my Cloudron homeserver and so an app (Mastodon) hosted on there went down for a few weeks. Not a big problem, just social media.

I'd like to restore from a 3 week old backup but then combine it with the most recent backup but am unsure of what to expect.

Here's a summary of the situation:

• mastodon app Data Directory hosted on volume (external USB harddisk) in /mastodon subdirectory

• app Data Directory was being updated nightly to separate backup harddisk

• last good app backup was 20231205-mastodon

• some issue with USB harddisk caused it to be unmounted

• tried to remount but didn't work

• app wasn't responding in cloudron so couldn't do nightly backup

• 3 weeks pass

• finally got home and remounted disk, this time remapped it as ext4 volume from within Cloudon volumes UI

When I tried to restart it it couldn't as disk mount location had changed (directory not empty message in Volumes UI)

• I decided to change the directory to mastodon-240108

App restarts with minimal data (app size went from ~25Gb down to ~2Gb) - as expected.

• app working. Some issues with user 2FA but these are now fixed.

• restarted nightly backups. Most recent 20240108-mastodon.

Question:
From the backup files I have available, I'd like to get back to the app state from 20231205-mastodon, and then preserve all activity from backup 20240108-mastodon to current.

Is this possible?

I don't know if the following will achieve this:

1. restore the 20231205-mastodon into the new mastodon app directory located in mastodon-240108
2. take backup
3. restore from 20240108-mastodon

Any suggestions or help greatly appreciated.

Thanks @BrutalBirdie. That might help. I think I may have the syntax wrong.

From your command in that post, in my case the box-TASKID.service is 10767, so I take it the correct syntax is:

watch -d -n 0.1 ls -lah /proc/$10767/fd/

is that right?

I'm getting the following issue

Every 0.1s: ls -lah /proc/10767/fd/                                                                        apps: Wed Aug 16 16:14:26 2023

ls: cannot access '/proc/10767/fd/': No such file or directory

Actually looking at the progress in the CLI:

{"percent":15,"message":"Backup - Uploading backup 60411M@5MBps 
{"percent":15,"message":"Backup - Uploading backup 60501M@9MBps
{"percent":15,"message":"Backup - Uploading backup 60648M@15MBps
{"percent":15,"message":"Backup - Uploading backup 60807M@16MBps
{"percent":15,"message":"Backup - Uploading backup 60890M@8MBps
{"percent":15,"message":"Backup - Uploading backup *61060M*@17MBps
{"percent":15,"message":"Backup - Uploading backup *61060M*@0MBps
{"percent":15,"message":"Backup - Uploading backup *61060M*@0MBps
...
{"percent":15,"message":"Backup - Uploading backup *219M*@5MBps
{"percent":15,"message":"Backup - Uploading backup *255M*@4MBps
{"percent":15,"message":"Backup - Uploading backup *278M*@2MBps

I expected those numbers before the M@xMBps would always increment. I thought they were showing how much had been backuped/updated already. Is this assumption incorrect?

(I did search previous messages but could not find any mention. Apologies if this was answered before)

I'm currently doing a backup of a heavily used app. It's been stuck at 15% for a number of hours. From the progress both the GUI indicator:

and CLI indicator:

Aug 16 15:49:53 box:tasks update 10767: {"percent":15,"message":"Backup - Uploading backup 55227M@10MBps (APP.DOMAIN.TLD)"}

it's unclear to me the units of the progress so far.

Is it 55227MB or Mb?

This combined with not being able to work out the total size means I'm unable to say if this has stalled, work out how long it'll take, etc.

Any help appreciated!

@nebulon said in App snapshot created but backup progress still pending:

The tooltip should be shown in the progressbar in the apps grid. By now the progress status in the app configure view is directly displayed. I guess the UI has changed a bit since this topic was created few years ago.

Aha it's shown in the apps grid, not the app page itself.

OK, thats still helpful. Thanks @nebulon.

@girish said in App snapshot created but backup progress still pending:

you can hover over the progress bar and it will show a tool tip as to what it is up to.

Hi @girish is this still true? I've not been able to get this tooltip working in Firefox Developer 114.0b3 (64-bit).

@nebulon That seemed to fix it. Thanks!

@girish said in Experience of first Penpot install by a UX Designer.:

@ei8fdb ah, I see why. The app has a nginx of it's own. nginx starts first and before penpot itself. So, Cloudron thinks app is running early (because of nginx running). I guess the healthcheck might need adjusting.

Ah so it's down to timing? As I said, I was clicking pretty quickly.

If you let me know when you do that adjustment I'd be happy to test it again.

@nebulon I just did a fresh install of a test instance. Here's a log of the ending of the install process: https://pastebin.ei8fdb.org/tuneseyizo.http

The Bad Gateway error did happen. But 2-3 seconds later the error was replaced by the correct login page.

To try and analyse it:

• In terms of my interactions, I was watching the install process in my dashboard, the instant the UI said "installed", I clicked on the app link, and I got the "Bad Gateway" error.

• In terms of logging, it happened somewhere around here:

Jun 14 18:23:20 [2023-06-14 17:23:20.970] I app.auth.oidc - 
hint="provider initialized", provider="oidc", method="discover", 
client-id="260ad1a1-a188-4f9d-879e-1f9657ce983c", 
client-secret="AAAAAAAA***************************************************************************************************************************", 
scopes="profile,email,openid", 
auth-uri="https://my.MYSITE.DOMAIN/openid/auth", 
user-uri="https://my.MYSITE.DOMAIN/openid/me", 
token-uri="https://my.MYSITE.DOMAIN/openid/token", 
roles-attr=nil, roles=nil
Jun 14 18:23:20 [2023-06-14 17:23:20.973] I app.metrics - 
action="initialize metrics"
Jun 14 18:23:20 [2023-06-14 17:23:20.984] I app.db - 
hint="initialize connection pool", name="main", 
uri="postgresql://postgresql/db260ad1a1a1884f9d879e1f9657ce983c", 
read-only=false, with-credentials=true, min-size=0, max-size=60

So, in terms of reproducing this error, the best I can suggest is click on the app link the second the installation process finishes.

Does anyone else have the time to try to reproduce my behaviours? Maybe it's simply, wait for a few seconds more.