I did a clean installation and restored a backup. Since than, everything is working fine. Thanks

@reachableceo thanks for the update!

Thank you all, I only thought the port forwarding of 80 was enough, Thank you for all the help for explanations

@3246 @girish , I am not sure (didn't try myself) but things suggested in this post may help ..

@girish thanks! That's indeed the fix..

It's fixed a bit differently. The purpose of the "profile lock" feature is that admins have control over the username/mail/display name. Letting the user choose the username would circumvent this. This more so applies because when you have multiple admins, maybe one admin forgot to set the username (but they had decided by policy that they won't allow users to set usernames themselves). Given that the profile lock is a 'dynamic' flag i.e can be set/unset at anytime, the fix we have put in place is: username can now be set already created users with no username. When profile lock is set, username is required when creating new users. This way it's not possible for an admin to create users without a username. When profile lock is set, invite links of users with no username show a message to contact admin. (This message will only be seen if you had users created with no username and set the profile lock later. So, it's kinda rare)

I solved the same symptoms by recreating Personal Access Token for Cloudron. [image: 1642094319690-screenshot-2022-01-13-at-6.18.06-pm.png] Read/Write access under Domains required. Mine just expired after default set 6 months.

@mark_ehm At what stage of onboarding are you? Do you have Cloudron installed and setup already? If not, our demo is a good place to start understanding what the product is/does. https://my.demo.cloudron.io/ (username and password is both cloudron). You would install something like Bookstack. You can login to bookstack with the same credentials (cloudron). You can add users in Cloudron dashboard and those users will get access to bookstack as well. If you already have Cloudron setup and working, then just ignore what I said and follow @jdaviescoates said

@robi yes!

(please read warning in above comment before following instructions below). If you want to make config file changes: docker exec -ti mysql /bin/bash Edit /run/mysql/my.cnf as needed supervisorctl restart mysql It is important to note that the changes to config files are not persistent across container restarts and server reboot. In general, we have not envisioned Cloudron as a development environment, but we do have many devs using it as such. So, if there is interest, maybe we need to improve things for such use cases.

This is possibly a bug, but in general we don't completely test as to what gets enabled/disabled when subscriptions expire. It's quite rare that people keep their servers around when subscription expires. But I will test things out when we cut a new release.

I figured it out. There was an issue in namecheap because after checking the ip adress in dnschecker it had two ip addresses: One was the server ip and the other from namecheap. Now there is only one available after chatting with them.

@meeksfamily06 Cloudron uses unbound and does not use systemd-resolved. You won't see any issues unless when trying to use the mail server (or maybe some apps like nextcloud might act strange when going to the apps section).

@ashkaptsov So, AppGyver's browser UI is going to make requests from the browser to n8n and baserow ? Can you tell us a bit more as to why would it want to do that? In general, adding CORS headers without knowing the app knowing about is a security hole.

@girish so I tried more things and my initial observation cannot be reproduced anymore, so I think I might have been misguided by persistent console logs here. Now I am not able to reproduce the double upload trigger at all anymore in chromium or firefox. @robi can you maybe explain in more detail how to reliably reproduce this? Do you have any additional extensions installed and can anyone else here see the issue?

Ah, thanks for reporting. I have opened https://git.cloudron.io/cloudron/box/-/issues/808 internally .

@joshrb91 looks like the server has nvm installed. Which then hijacked the system node. # which node /root/.nvm/versions/node/v14.17.6/bin/node The error is not very obvious because systemctl status box says running but it's not really running. journalctl -u box -fa shows that it's crashing constantly. The reason for that is the box code runs as yellowtent user.. # su - yellowtent $ node --version v12.18.4 $ which node /usr/local/bin/node It seems there is another node installed in /usr/local/bin also. I renamed that directory to bin-old for now and it brings things up. Generally, we don't support installing external software on Cloudron. Was there any reason you installed nvm on the server? (was it to install cloudron CLI? if so, that's meant to be installed on your PC/laptop and not on the server).

@jtippetts It's hard to tell what is happening. Can you try this: Reboot the server Please check if the IP address of server matches the DNS entries. Maybe the IP address was changed during the process of opening up email ports? Once rebooted, please check the output of systemctl status box and systemctl status nginx. If they are both running, the dashboard should be reachable.

Adding a redirection made things work.