Cloudron 5 released

[girish]

What we initially planned as Cloudron 4.5 is now Cloudron 5. We decided to bump the major version since it got a lot of changes to justify the bump.

New installations will already get Cloudron 5. Existing installations will get an update in a week or two. We roll out updates conservatively as always, if you want it immediately, please just send a mail to support@cloudron.io and we will make it available.

I am preparing the blog post which will have the screenshots for the features.

Changes:

• Add user roles - owner, admin, user manager and user

• Show backup disk usage in graphs

• Mail server event log and mailbox usage information

• Spam auto-learning - the spam solution is now per mailbox. The mail server will now periodically auto learn spam and ham from the mailbox with no user intervention.

• CPU resource allocation - We have implemented CPU Shared

• Whitelabel the footer/404

• App specific passwords

• Import backups UI - this lets you easily move apps across Cloudrons

• Display timestamps in browser timezone in the UI

• linode: add object storage backend

• Enhancements in the built-in LDAP to better support SOGo (more on this in a separate post).

There's a lot of other bug fixes as well. You can read through the commits for that

[tamayers]

Awesome news! Great work to you and the team!!

[jdaviescoates]

Fantastic news, thanks!

[will]

Great work, you guys have been quiet, that usually means furious work is being done behind the scenes.

[vjvanjungg]

thank u! I just tried v.5 on the demo page yesterday, great job

[will]

@vjvanjungg same, so fa so good!

[necrevistonnezr]

Works great so far!
I have no SSD / HDD graphs, though:

df -h output:

Dateisystem                 Größe Benutzt Verf. Verw% Eingehängt auf
udev                         3,8G       0  3,8G    0% /dev
tmpfs                        765M    3,2M  762M    1% /run
/dev/mapper/beebox--vg-root  455G    328G  104G   76% /
tmpfs                        3,8G    8,0K  3,8G    1% /dev/shm
tmpfs                        5,0M       0  5,0M    0% /run/lock
tmpfs                        3,8G       0  3,8G    0% /sys/fs/cgroup
/dev/loop0                   9,2M    9,2M     0  100% /snap/canonical-livepatch/93
/dev/loop1                    92M     92M     0  100% /snap/core/8689
/dev/loop2                    55M     55M     0  100% /snap/core18/1668
/dev/loop3                   9,2M    9,2M     0  100% /snap/canonical-livepatch/94
/dev/loop4                    92M     92M     0  100% /snap/core/8592
/dev/sdb1                    472M    145M  303M   33% /boot
/dev/sda1                    1,8T    1,5T  306G   83% /media/SG2TB

[d19dotca]

@necrevistonnezr According to the demo, the disk has moved to the System page (which honestly this memory graph should too IMO, it's system info, don't need a dedicated page for just memory and disk usage IMO).

[necrevistonnezr]

Thanks, I'm blind.

In that screen, I just see a spinning circle....

[girish]

@necrevistonnezr Anything errors in the browser console? Looks like we broke something. The idea was to move the graphs into two parts in the next release - app memory graphs will move into the apps itself and the summary will move into system info.

[girish]

@d19dotca Yes, that's the plan we wanted to get what we had released, since we were accumulating too many changes.

[necrevistonnezr]

@girish said in Cloudron 5 released:

@necrevistonnezr Anything errors in the browser console? Looks like we broke something. The idea was to move the graphs into two parts in the next release - app memory graphs will move into the apps itself and the summary will move into system info.

From the browser console (Firefox):

Content Security Policy: Duplicate source 'unsafe-inline' is ignored
Content Security Policy: Invalid host 'unsafe-inline' cannot be processed
Content Security Policy: Duplicate source 'unsafe-inline' is ignored
Content Security Policy: Invalid host 'unsafe-inline' cannot be processed
This page uses the non-standard property "zoom". Instead, calc() should be used in the corresponding property values or "transform" together with "transform-origin: 0 0".
SingleFile is hooking the requestAnimationFrame and cancelAnimationFrame functions to load deferred images in background tabs. content-hooks-frames-web.js:231:9

[girish]

We have a regression that we cannot send email from Windows/Outlook clients. This is because these clients do not set a EHLO name that is a fqdn. Cloudron 5 enabled a check for validating host EHLO hostnames which is causing this problem. To fix this,

1. SSH into Cloudron server
2. docker exec -ti mail /bin/bash
   2.1 Edit /run/haraka/config/helo.checks.ini
   2.2 Change valid_hostname=true to valid_hostname=false
   2.3 supervisorctl restart haraka
3. Important: the changes are not persisted across mail container and server restarts. So you have to do the above manually for now, until we fix the issue. We should have a patch release in a day or so.

[msbt]

@girish can we check if mails didn't get delivered in the new email view?

[girish]

@msbt The eventlog is a work in progress, currently it's a bit raw. But it does show that email got delivered. look for the outbound icon <- and then if you click it will say 'delivered' in the expanded logs. If it didn't get delivered, you will see that it's retrying.

[Trankery]

Did you remove the reboot server option?

[msbt]

@Trankery it will come back

[Trankery]

@msbt Thanks!

[girish]

It seems the EHLO validation has another bug wrt dynamic host names. To fix this:

1. SSH into Cloudron server
2. docker exec -ti mail /bin/bash
   2.1 Edit /run/haraka/config/helo.checks.ini
   2.2 Change dynamic=true" in the [reject] section to dynamic=false
   2.3 supervisorctl restart haraka

Important: the changes are not persisted across mail container and server restarts. So you have to do the above manually for now, until we fix the issue. We should have a patch release in a day or so.

[girish]

The blog is out with screenshots! - https://cloudron.io/blog/2020-03-26-cloudron-5.html