Cloudron Forum

Thank you both @robi and @nebulon

Got it working.

Here are the steps in case if anybody wishes to replicate the process.
It would help a lot to the new Evcc users if the package includes the mentioned changes already.

Steps:

1. Installed Evcc
2. Click: File Manager
3. Click: New > New File
4. Name it this text: evcc.yaml
5. Click: Restart App
6. Refresh the web browser window. A new file will be created named: evcc.db
7. Open the file: evcc.yaml
   There is already a text like this inside:

network:
  schema: http
  port: 7070
  host: evcc.demo.cloudron.io
database:
  type: sqlite
  dsn: /app/data/evcc.db

Click enter 2 times to be able to write to a new line bellow

8. Copy-paste the text from here:
   https://docs.evcc.io/en/docs/installation/configuration#manual

This is the text:

## minimal configuration example

site:
  title: Home # display name for UI
  meters:
    grid: my_grid
    pv:
      - my_pv
    battery:
      - my_battery

# define your loadpoints according your needs
# see https://docs.evcc.io/en/docs/reference/configuration/loadpoints
loadpoints:
  - title: Garage # display name for UI
    charger: my_charger # charger
    vehicle: my_car # default vehicle

# meter definitions
# name can be freely chosen and is used as reference when assigning meters to site and loadpoints
# for documentation see https://docs.evcc.io/docs/devices/meters
meters:
  # replace with your real grid meter
  - name: my_grid
    type: template
    template: demo-meter
    usage: grid
    power: -1000 # 1 kW feed-in
  # replace with your real solar system
  - name: my_pv
    type: template
    template: demo-meter
    usage: pv
    power: 4000 # 4 kW production
  # replace with your real battery
  - name: my_battery
    type: template
    template: demo-battery
    usage: battery
    power: -1000 # 1 kW battery charging
    soc: 50 # 50 % state of charge

# replace with your real charger
# see https://docs.evcc.io/docs/devices/chargers
chargers:
  - name: my_charger
    type: template
    template: demo-charger
    status: C # charging
    power: 2000 # 2 kW charging power
    enabled: true # optional

# replace with your real vehicle (optional)
# see https://docs.evcc.io/docs/devices/vehicles
vehicles:
  - name: my_car
    type: template
    template: offline
    title: blue e-Golf
    capacity: 50 # in kWh

# enter your real grid tariff and feed-in price
# see https://docs.evcc.io/docs/tariffs
tariffs:
  currency: EUR
  grid:
    type: fixed
    price: 0.29 # EUR/kWh
  feedin:
    type: fixed
    price: 0.10 # EUR/kWh

9. Click: Save + Restart App

10. Click on Evcc (URL)

11. There we can see Demo data for: blue e-Golf - charging

Add a new Admin password
Click: Configuration + Add the password again
Click to activate the Experimental features

After this hopefully, the changes can be performed directly inside Evcc via the web browser UI.
If not, the evcc.yaml file will have to be edited manually,
because these terminal commands still aren't working: evcc detect + evcc configure

Hope that was useful

Hi everybody

It's logical that the following scenario would occur:

1. As more Community apps are added
2. Those that have been marked as very stable
3. Can be processed by the Cloudron team and added in the official Cloudron Catalog

---

In that scenario, the people that have been using a Community App
(examples: Hermes, Fizzy, Garage S3, Ente, Browserless),
would have 2 choices:

1. Stay on the Community version (if the maintainer decides to keep maintaining it)

2. Switch to get the new updates from the Cloudron source

• This post is about that.
• It would be awesome if the Cloudron team could (some day) release an update which includes the option to switch from the Community source, to the Cloudron source.
• This way, a migration is not needed and the transition goes smooth.

---

This way, the Cloudron Catalog can grow much faster while with max stability.

Hope that was useful

Thanks @girish
Us, the people that install softwares through Cloudron (Cloudron catalog + Community apps) worry only about 2 things:

• That each app is updated regularly
• Stability & Performance

Who packages and maintains it, doesn't matter much.
If a community maintainer does a good job, great. We would gladly donate to support his/her work.

3 scenarios exist:

1. If the Cloudron team planned on adding couple Apps in the Catalog (which a community maintainer was faster to release on the Community list), and now the App isn't added on the list because somebody was faster.
   Yes, some maintainers do a great job. But not all are up to the gold standard that the Cloudron team has. Most of those people have regular jobs, so they can't dedicate the additional time.

2. If a Community maintainer, stops maintaining an App. And if the Cloudron team picks to add the same App on the Catalog.
   All those hundreds/thousands of people that initially used the Community version, would need an streamlined switch to the Cloudron version. Ideally without a full migration.
   This is where it would be very useful if some day in the future the Cloudron team could create an option in the Community App dashboard, to just click to change the .json URL - so that the App continues to get the next updates from the Cloudron source.

3. A Community maintainer want's to help the community and the Cloudron team by first maintaining an App himself, but with the purpose in the future to delegate the same App to the Cloudron team, so that he can redirect his focus on preparing new Apps.

Thank you very much
@robi @joseph @crazybrad and @p44

The settings are correct because we are also using Backblaze for the backups.

We tried all your recommendations and nothing worked.

Has ANYBODY successfully performed the Typebot-BackblazeB2 connection?

If yes, please share the settings. We will perform the tests on our side also. Then we will send the detailed steps for the Cloudron developers to add in the Typebot documentation (if they wish) for everybody to benefit

Thank you @andreasdueren Andreas for maintaining Hermes. It's a fascinating software.

We are having some difficulty reaching the /kanban/ dashboard.
If possible please forward the steps that you have made to achieve that outcome.

These are our steps:

1. Open a new web browser (allowed third-party cookies)
2. Go at: https://url.com/kanban/
3. Enter the Cloudron user credentials
4. Redirected at: https://url.com/login?redirect=/kanban/
5. Click again: Log in
6. The web browser window refreshes and nothing happens.

From our research, it looks like the Cloudron login system doesn't allow the redirection successfully.

Congratz Team Cloudron
Is there an approximate timeline for 9.1?

The Custom app build & deploy + Community packages are something that the majority of us have been looking forward to for a long time

The latest version of Cloudron 7.7.2
made all newly installed Uptime Kuma's to be unavailable to be loaded upon installation.

I even tried to install older versions of Uptime Kuma. Again the same result.

Then I did the installation on Cloudron 7.7.1 and it works.

If possible please fix this bug before pushing 7.7.2 in the Stable Updates section.
Thanks

@andreasdueren Thank you very much. I confirm that OpenClaw is now running without errors.

Thank you @james

We have performed the steps and it's working

Here are the steps in detail (for noobs like me to replicate it):

1. Log in Cloudron + select EspoCRM + Click on the Repair tab: Enable Recovery Mode
2. Click: Terminal
3. Copy paste this:
   nano +704 application/Espo/Core/Utils/File/Manager.php
4. This is the original line:
   if (is_dir($dirPath) && is_writable($dirPath)) {
5. Change it to this:
   if (is_dir($dirPath) && is_writable($dirPath) && is_link($dirPath) === false) {
6. On the keyboard click: CTRL + X
   Then click to confirm: Y
7. Copy paste this:
   mkdir -p /run/espocrm/cache
   mkdir -p /run/espocrm/logs
   php command.php extension -u --name="Advanced Pack"
8. Close the Terminal window
9. Click to Disable the Recovery mode
   there is no need to Restart EspoCRM
10. Log in EspoCRM as admin >>> extensions

• 1 more step to update the Extension:
  Since the old Extension version was uninstalled, now we upload the new version + click Install

@james We noticed that the change in the Manager.php file is not permanent.
So this means that we would have to go through all these steps each time an Extension update is required.

Business-wise, this will reduce the client-retention rate of the Cloudron userbase. Especially those that aren't tech experienced.

Ideally the extension updates would be performed by the EspoCRM admin just uploading the new version inside EspoCRM (UI) and it replaces the old version.
Is it possible for the Cloudron team some day to push an update to the EspoCRM package that applies this change permanently?

@girish Thank you very much.

We run the tests again by reinstalling it again.
It seems that the problem originated when we installed multiple Apps and then UptimeKuma somehow broke. So we installed them one by one, and now it's ok.

@joseph Thank you very much. That fixed the issue

the post can be marked as solved

Hi

When we try to edit the config files using the File Manager for any app this happens:
(Example with Matrix)

1. Edit file: homeserver.yaml (owner: cloudron)
2. Click: Save + Close
3. In the folder there is a new file: homeserver.yaml.1 (owner: root)
4. Click: Restart App
5. The new file has changed owner: homeserver.yaml.1 (owner: cloudron)
6. And there aren't any changes performed inside: homeserver.yaml (owner: cloudron)
7. The changes aren't applied in the app

Is it possibly something that we are doing wrong, or is it just a temporary bug?

Bug report for version: 9.2.0

The users permissions is missing when we need to add a new User + then modify existing users permissions

With the recent changes to MinIO, we are forced to use another solution for the Typebot storage.

So we took a look at BackblazeB2.
Backblaze is very cost-effective 1TB=$6/m so it's a good solution.

We tried the following configuration:

1. Inside Backblaze
   Activated the Bucket
   Set CORS as receive from all
   Left the Lifecycle rules the same
   In the settings tested both Private + Public

2. In the Typebot backend file added the standard: (for this specific bucket)

export S3_ACCESS_KEY=XXXXXXXXXXXXXXXXXXXXXXXXXXXXXX
export S3_SECRET_KEY=XXXXXXXXXXXXXXXXXXXXXXXXXXXXXX
export S3_BUCKET=typebotstorage
export S3_ENDPOINT=s3.us-east-005.backblazeb2.com
export S3_SSL=true
export S3_REGION=us-east-005
+
It didn't work.

Made tests with:
export STORAGE_TYPE=s3
export S3_REGION=US
export S3_REGION=us
export S3_PORT=443
+
Also, no results.

And yes, we performed an App restart per each individual change.

Does anyone have an idea how to make it work?

Thank you @james

We performed these steps and it was successful:

1. Add these lines in env.sh
   export NODES_EXCLUDE="["ExecuteCommand", "LocalFileTrigger"]"

2. Restarted N8N

3. Performed the test using these steps:

• (+) - Execute command
• Execute Once: ON
• Command: ffmpeg -version
• Test step
• Check the OUTPUT - ok (screenshot attached)

1. We installed Evcc

2. We see this notification:
   Run evcc detect or evcc configure for initial setup. Then restart the app.

3. We went in the EVCC - Terminal and have run both commands and this is the result:

Connecting...
root@448da2a9-736b-4cc3-9370-1e160d939387:/app/code# evcc detect
This account is currently not available.
root@448da2a9-736b-4cc3-9370-1e160d939387:/app/code# evcc configure
This account is currently not available.
root@448da2a9-736b-4cc3-9370-1e160d939387:/app/code#

Since there isn't any post in this category (except for the Package Updates post), has anybody configured Evcc and confirmed it's functionality?

@andreasdueren Thank you, much appreciated.
Soon I'll test Zai, Ollama and Mattermost

Here is something possibly useful.
On our OpenClaw installation via DokPloy, we installed SearXNG on the same VPS.
And we were able to configure it to be able to perform the web searches directly via it. So zero API cost towards Brave/Perplexity.

The Docker Compose value that we added was this one:
SEARXNG_URL=

Perhaps somebody else would wish to apply the same configuration. It would be beneficial if it's also added. Thanks again

Thank you @robi for the fast response. We tested that and the result is the same

Thank you @andreasdueren for packaging OpenClaw. Great work!

Due to it's high customizability, it requires more env options to be added.
From my research, here is a list of some of the seem to be currently missing in the first version of the package (hope it's useful)
Note: Some may be redundant, so not all would need to be added.

ZAI_API_KEY=
GROQ_API_KEY=
MISTRAL_API_KEY=
CEREBRAS_API_KEY=
VENICE_API_KEY=
MOONSHOT_API_KEY=
KIMI_API_KEY=
MINIMAX_API_KEY=
AI_GATEWAY_API_KEY=
OPENCODE_API_KEY=
SYNTHETIC_API_KEY=
COPILOT_GITHUB_TOKEN=
XIAOMI_API_KEY=
OPENCLAW_PRIMARY_MODEL=
OLLAMA_API_KEY=
OLLAMA_BASE_URL=
BROWSER_CDP_URL=
BROWSER_DEFAULT_PROFILE=
BROWSER_EVALUATE_ENABLED=
BROWSER_SNAPSHOT_MODE=
BROWSER_REMOTE_TIMEOUT_MS=
BROWSER_REMOTE_HANDSHAKE_TIMEOUT_MS=
HOOKS_ENABLED=
HOOKS_PATH=
WHATSAPP_ENABLED=
MATTERMOST_BOT_TOKEN=
MATTERMOST_URL=
ZALO_BOT_TOKEN=
OPENCLAW_TWITCH_ACCESS_TOKEN=
SLACK_SIGNING_SECRET=
LINE_CHANNEL_SECRET=
LINE_CHANNEL_ACCESS_TOKEN=
DISCORD_CLIENT_ID=
TEAMS_APP_ID=
TEAMS_APP_PASSWORD=

@robi In the GitHub repository the breaking changes are listed
https://github.com/openclaw/openclaw/releases

Example (from our current version 2026.3.23 to the current one 2026.4.11):
openclaw 2026.4.5
openclaw 2026.4.2
openclaw 2026.3.31
openclaw 2026.3.28
openclaw 2026.3.24

Plus regular upgrades to the system functionality, for example how the Videos are being generated.
But this is due to the setup that each client has performed before that.
For example, to be able to generate Videos via Gemini, we had to install custom skills and scripts. So when the update is applied that doesn't have that requirement, either that functionality will break the current process, or we will just need to re-adjust the configuration.

Many people have reported that they are performing the breakage fixes just by connecting Claude Code via SSH and letting it do it's "magic".