For the record, I hit that issue too. /boot got full. However, I could not fix it with the solution at https://forum.cloudron.io/topic/1310/disk-filled-up-with-kernel-update-artifacts/2 . I had to follow https://askubuntu.com/a/430944 (slightly adapted)

@nebulon said in Disk filled up with kernel update artifacts:

rm -rf /var/tmp/mkinitramfs*

I had to instead do:

cd /var/tmp find . -type f -delete -print

This was running for like 5 hours! (and still running)

In another server instance hit by https://forum.cloudron.io/topic/1382/error-for-multiple-apps/12, recreating docker containers did not help. I even disabled apparmor altogether using sudo systemctl stop apparmor and sudo systemctl disable apparmor. We tried server restart as well and that didn't help either.

After disabling apparmor, I saw that mysqld would just quit without showing any error message or log output 😞 As a final thing, I just ended up moving platformdata/mysql to mysql-old and recreated containers again. That made mysql come up and then we restored all the apps.

Still a bit crazy/worrying that we hit this bug atleast twice now and there is no clear resolution.

For others reading this, currently the process of giving build service access is not automated. If you need access, please send us an email to support@cloudron.io.

@rtyu1120 for hackmd there is already some progress https://forum.cloudron.io/topic/1222/hackmd-fast-and-flexible-real-time-collaborative-markdown/10

@girish I also realized that the mysqldump for the TT-RSS app is huge (679 MB in my case). Could you implement some compession in the database dumps (and use PostgreSQL as recommended by the author)?

@uiharu Ah, thanks!

Ah my bad, for whatever reason I deleted the DNS entry and didn't check what was up, all my fault :s

@girish working. thanks!

Thanks again! Everything is up and running now!

@scooke Right, so the newapp.app@domain.com is the default allocated mailbox. If you want to make the app use a different email address, you have to do this https://cloudron.io/documentation/email/#changing-the-from-address-of-an-app . It's geeky, we are looking into fixing this in the next release since this is often requested.

Awesome. Thanks!

@fbartels @girish Thank you for the support.

Solution:

use a SMTP mail solution sendmail (and the mail()) are not usable in current state PHPMailer could be used as an alternative

@girish Okay, workaround works ! 🙂

@nebulon said in Nextcloud: Empty Trashbin intervals:

trashbin_retention_obligation

@nebulon thank you very much! Missed that the config.php is symlinked to /app/data and therefore it is editable 👍🏼

Upon some investigating, did find a polls utility like doodle in nextcloud. It allows you to share it with users but also share a public link. Pols can be text based or created using a calendar. Persons responding can answer a yes,no and maybe on whatever options are offered.
https://github.com/nextcloud/polls

@TheNils For new domains, the NS record sometimes takes to propagate. That error message means that it's not propagated to the Cloudron server yet. You can try this:

SSH into the Cloudron host -t NS mydomaine.net (This should not display NXDOMAIN) If the above works, it should work on Cloudron as well. If not, can try sudo systemctl restart unbound and then try adding the domain again?

We will improve the error message to be more descriptive (https://git.cloudron.io/cloudron/box/issues/548).

@nebulon Thank you 🙂

Hello! Sorry - yes inbound 80 is open. results of journalctl:

No journal files were found.

Normal users can create tokens but they don't have access to any call other than the /api/v1/profile/* routes. Internally, each token has a list of "scopes" (oauth scopes) which indicate what API can be allowed. For normal users, this scope is only the profile scope. For admin users, it includes all the other API calls.

@girish The update worked! The first steps didn't work for me, but the apt update did the trick. Thank you!

@girish Thanks for the response. I have wiped the server and did a restore, so I can't do what you suggest atm.

There is/was plenty of space (42gb used out of 490), and ram (1gb out of 6gb).

I was/am running 2.0.1.

I'll check the event log size once the problem comes in again and will report here.

@x1tan I have made the page publicly accessible now. For some reason it had a wrong setting there.

I have updated the ciphers now according to mozilla's config generator. The commit is https://git.cloudron.io/cloudron/box/commit/ddaa52163bf3844b36d6c29fdffb5db3e0b3f5d0

For the CSP settings, this indeed cannot properly be done on a platform level, as apps require differently strict settings there and have to provide this on their own, so this should ideally be fixed in each app upstream.

@girish I have a 238 ko iCal file.
I've split it in twice and I went from 250 importation err from 1 err.
The problem should come from that, thanks!

I was at the beginning thinking about it (splitting the file) but after that, I was focusing on RateLimit - wrong way! Also, I advise people which want to import a big file to increase drastically the mem allowance of the app (2Gb in my case)

Yep! I am on a Lightsail instance and simply needed to open up port 25 for traffic. Thanks!

It works. Thanks.

Hi there,

running on bare metal is totally fine. For the setup within your network behind the router, please make sure that at least port 80 and 443 are forwarded. Port 80 is required to obtain LetsEncrypt SSL certificates. Otherwise please check the logs with journalctl -u box when performing the dns setup on your Cloudron, this should show for which IP it is waiting for the DNS records to be in-sync. Possibly it is checking for the wrong (private) IP.

Further when using Cloudrflare, please note that currently Cloudron does not support installing apps that are proxied via Cloudflare. Cloudflare backend only sets up the DNS via Cloudflare API and expects website traffic to be unproxied.