VestaCP 18.10.1 Compatibility ETA & Bad gateway nginx after updating.

[plutocrat]

Seems somehow the problem is to do with upgrading from Ubuntu 16 to 18. It does actually work now on a fresh install of 18 (although it didn't for a long time). Would a fresh install be an option for you?

[p4r4norm4l]

Hey I upgraded from 18.04 and not from 16.
According to vestaCP website it's not compatible with 18.10.1 yet.

I could do a full backup and reinstall everything, I did the update because wordpress is now compatible with php 7.3 and speeds up the websites by 9%.

I'll see maybe i'll just wait and use Vesta as CLI I was able to do the Backup that way so i can probably create new sites that way too.

I'll contact VestaCP about the issue and ask them if there is a patch coming out soon to support 18.10.1.

[imperio]

vesta is not compatible with 18.10, but we have a plans to support it after 0.9.8-24

[p4r4norm4l]

Do you have an ETA on that, Is there a quick fix or workaround for now ?

[plutocrat]

Oh, right, so it installed OK On 18.04, but you upgraded to 18.10 to get PHP 7.3? Well you know, its pretty easy to bump PHP up to 7.3 on an existing install. In retrospect that probably would have been an easier path! :-)
For the record, it goes something like this

Code: Select all

# Add repos and upgrade
apt install software-properties-common python-software-properties
LC_ALL=C.UTF-8 add-apt-repository ppa:ondrej/php
LC_ALL=C.UTF-8 add-apt-repository ppa:ondrej/apache2
apt update
apt install libapache2-mod-php7.3 php7.3 php7.3-cgi php7.3-cli php7.3-common php7.3-curl php7.3-gd php7.3-imap php7.3-intl php7.3-json php7.3-ldap php7.3-mbstring php7.3-mysql php7.3-opcache php7.3-pspell php7.3-readline php7.3-soap php7.3-xml 
php --version

# Fix apache
a2dismod php7.0 
a2enmod php7.3
systemctl restart apache2

[p4r4norm4l]

I will try this on a test server since the other one is a live server. :(
Should of done this the other way around lol.

[p4r4norm4l]

vesta is not compatible with 18.10, but we have a plans to support it after 0.9.8-24

Did the update and I still cannot conect to the admin panel. Did you guys do the switch from libcurl3 to 4 yet ?

[imperio]

vesta is not compatible with 18.10, but we have a plans to support it after 0.9.8-24

Did the update and I still cannot conect to the admin panel. Did you guys do the switch from libcurl3 to 4 yet ?

Can you write more about this problem?

[dtel]

found a workaround by copying /usr/lib/x86_64-linux-gnu/libcurl.so.4.4.0 from another working system and modifying the init script by adding LD_PRELOAD before starting command:

start_nginx() {
LD_PRELOAD=/usr/lib/x86_64-linux-gnu/libcurl.so.4.4.0 start-stop-daemon --start --quiet --pidfile $NGINX_PID \
--retry 5 --exec $NGINX_DAEMON --oknodo
}

start_php() {
LD_PRELOAD=/usr/lib/x86_64-linux-gnu/libcurl.so.4.4.0 start-stop-daemon --start --quiet --pidfile $PHP_PID \
--retry 5 --exec $PHP_DAEMON --oknodo
}

stop_nginx() {
LD_PRELOAD=/usr/lib/x86_64-linux-gnu/libcurl.so.4.4.0 start-stop-daemon --stop --quiet --pidfile $NGINX_PID \
--retry 5 --oknodo --exec $NGINX_DAEMON
}

stop_php() {
LD_PRELOAD=/usr/lib/x86_64-linux-gnu/libcurl.so.4.4.0 start-stop-daemon --stop --quiet --pidfile $PHP_PID \
--retry 5 --oknodo --exec $PHP_DAEMON
}


So far I succeded to have the backend functional again, the cron jobs using the new curl (4.5.0) are working and everything seems to be ok.

[drsdre]

Given that Ubuntu 16.04/18.04 uses systemd, you can also overrule the environment by creating a service config:

Code: Select all

systemctl edit vesta

Add to the new config file:

Code: Select all

[Service]
Environment="LD_PRELOAD=/usr/lib/x86_64-linux-gnu/libcurl.so.4.4.0"

Don't forget to update the daemon:

Code: Select all

systemctl daemon-reload

And restart the service:

Code: Select all

service restart vesta