• If you are still using CentOS 7.9, it's time to convert to Alma 8 with the free centos2alma tool by Plesk or Plesk Migrator. Please let us know your experiences or concerns in this thread:
    CentOS2Alma discussion

Issue Plesk Onyx, Horde and the Backup of 12.5

F!rsT S0uL

New Pleskian
Hi there,

So here i go:

i've installed a fresh Ubuntu 16.04 LTS, then i installed Plesk Onyx with my componetents.
Then i take the Plesk Backup of 12.5 from my FTP an restored it in Onyx. At the first sight it seems to be okay... then the following:

1) All Subdomains have the Error 503. Why? No idea. I found nothing in the logs or something.

2) I use Horde as webmail.
2a) ActiveSync doesn't work anymore. (Look: http://www.stiftnet.de/activesync-mit-horde-einsetzen/)

2b) Redirect on HTTPs doesn't work anymore. The files have changed (Look: https://der-linux-admin.de/2014/08/plesk-webmail-per-redirect-zu-https-umleiten/)

2c) That problem/bug is not solved and the workaround doesn't work anymore (Look: https://oli.new-lan.de/2015/02/plesk-12-mit-horde-5-2-caldav-webdav-authentication-failed/)

2d) In the Plesk Backup is NOT the Horde SQL DB. There are the calander, contacts etc. saved... I must restore a backup of the DB. Why?! Plesk should backup that!

3) Is SSLv3 enabled? So i must disable it manually?!

At this point i give up and restored a snapshot of my machine. The downtime was too long. I think there ist more things, that doesn't work anymore.

Any ideas?
Thanks!
 
According to your issues with horde:
The webmail config's layout was changed
That means your customisations do not work any more, you have to adapt it again.
And that is expected, because we cannot implement any new feature without touch of webserver configs.
For example, in webmail SSL certificates are supported per domain now - this the reason we split server-wide configs into separate domains.

By the way restoration of a backup created in a previous version is not a preferable for migration.
You can try Migration Manager component.
 
Hi F!rsT S0uL,

All Subdomains have the Error 503. Why? No idea. I found nothing in the logs or something.
Too less informations for decent investigations. In most cases, "503" errors appear due to a non-working service, or a misconfiguration ( apache2, nginx, php-fpm... ).

2) I use Horde as webmail.
2a) ActiveSync doesn't work anymore. (Look: http://www.stiftnet.de/activesync-mit-horde-einsetzen/)

2b) Redirect on HTTPs doesn't work anymore. The files have changed (Look: https://der-linux-admin.de/2014/08/plesk-webmail-per-redirect-zu-https-umleiten/)

2c) That problem/bug is not solved and the workaround doesn't work anymore (Look: https://oli.new-lan.de/2015/02/plesk-12-mit-horde-5-2-caldav-webdav-authentication-failed/)
Already answered by @EugeneKazakov

2d) In the Plesk Backup is NOT the Horde SQL DB. There are the calander, contacts etc. saved... I must restore a backup of the DB. Why?! Plesk should backup that!
DAILY "mysql.daily.dump.X.gz" - dumps ( databases: psa, horde, mysql ), where the lowest number ( defined as "X" in my example ) is always the latest daily dump, can be restored very fast and easy, with the commands:

Code:
gunzip /var/lib/psa/dumps/mysql.daily.dump.X.gz
mysql -uadmin -p`cat /etc/psa/.psa.shadow` -e"DROP DATABASE horde;"
mysql -f -uadmin -p`cat /etc/psa/.psa.shadow` < /var/lib/psa/dumps/mysql.daily.dump.X


3) Is SSLv3 enabled? So i must disable it manually?!
You might be interested in reading:

 
According to your issues with horde:
The webmail config's layout was changed
That means your customisations do not work any more, you have to adapt it again.
And that is expected, because we cannot implement any new feature without touch of webserver configs.
For example, in webmail SSL certificates are supported per domain now - this the reason we split server-wide configs into separate domains.

By the way restoration of a backup created in a previous version is not a preferable for migration.
You can try Migration Manager component.

Yes. I know. I did the customisations again for ActiveSync. But it doens't work... After i changed Apache from Event to Prefork and activated PHP7 Module, i got the error "Wrong Password" on my devices. Maybe this is the problem with the restoring of the old version...
Yep. I saw the function with webmail SSL... But i use a global Multidomain cert and i can't chose this in the config. So i uploaded the cert to the domain, changed the config, ticked in that a redirect on all domains and http requests and... it doensn't work. No change.

I will check the Migration Extension from Plesk. Then i mus run two servers. Is there another way with a backup and only one server?!

Hi F!rsT S0uL,


Too less informations for decent investigations. In most cases, "503" errors appear due to a non-working service, or a misconfiguration ( apache2, nginx, php-fpm... ).


Already answered by @EugeneKazakov


DAILY "mysql.daily.dump.X.gz" - dumps ( databases: psa, horde, mysql ), where the lowest number ( defined as "X" in my example ) is always the latest daily dump, can be restored very fast and easy, with the commands:

Code:
gunzip /var/lib/psa/dumps/mysql.daily.dump.X.gz
mysql -uadmin -p`cat /etc/psa/.psa.shadow` -e"DROP DATABASE horde;"
mysql -f -uadmin -p`cat /etc/psa/.psa.shadow` < /var/lib/psa/dumps/mysql.daily.dump.X



You might be interested in reading:


Error 503: Sorry... All services are working and i can't find anything in the logs...

SQL Dumps: are okay... but i thought they are in the backup of the "Backup Manager"

SSLv3: Great! Thanks!
 
Back
Top