@deiwee I guess you would want to run it as www-data user though to not mess with file permissions. Did you get any error messages in the log? Of course after app restart it will always take at most a minute to kick the cron according to schedule.

Have you tried installing it on your own? Please see Lamp App - Installing custom PHP extensions 🙂

@jlx89 there is no internal cache. Can you just do a plain curl -v https://clonedapp.domain.com and see what is happening? Could it be that you have the domain hardcoded in your LAMP app and it's redirecting you to that location? This happens fairly often when cloning WP with misconfigured plugins.

@carbonbee don't think it matters but I will push an update to phpMyAdmin 5.1.0 - https://www.phpmyadmin.net/news/2021/2/24/phpmyadmin-510-released/

@girish Hah this actually helps me with the Humhub app. Thanks! My solution was to use sed to delete line 312 in /etc/php/7.4/cli/php.ini lmao. Jank to the extreme.

@nj I have fixed this in the new package. However, it will work only for new installation. For existing install, you have to edit /app/data/apache/app.conf and add the following lines (You can add it after the DocumentRoot /app/data/public line):

LogFormat "%{X-Forwarded-For}i %l %u %t \"%r\" %>s %b \"%{Referer}i\" \"%{User-Agent}i\"" proxy CustomLog "|/bin/cat" proxy ErrorLog "|/bin/cat"

After that restart the app.

@shai you have to build a custom app, please see this guide https://docs.cloudron.io/guides/multiple-databases-lamp/

@girish

But a random web search

+1 for random web searches!

led me to where someone suggests that "I had this issue in drush command line, when there was an error parsing the settings.php file

That was it! A missing comma in the databases array. Thanks so much!

@ccfu I am not quite sure anymore why we are explicitly overriding that header, even if it is passed from the upstream server (ie the app in this case)

I've pushed a change for next release which will allow overriding it now https://git.cloudron.io/cloudron/box/-/commit/70f2337b09f321c1252fd0e17754e5ffbb1080d9

Does anyone know if this may have problematic side-effects security wise?

@hardwarehulp Are you referring to the LAMP app? It looks like that and I have preemptively moved this request to that category.

@timconsidine fixed with https://git.cloudron.io/cloudron/docker-base-image/-/commit/c3431324eaddac83122e3e63893c7c6a8e1ee17d but of course only part of the next base image release then.

Indeed, the current composer comes from ubuntu 20. If you want a different version, you can always download the phar from the composer website and run it as @murgero suggested.

@girish surveyjs looks great.
It can certainly help with some projects/use cases.
But think I will pursue Budibase as the more complete design/host/collect responses solutions.
Got it working (easily after stupid error on my side) and will look at putting on a VPS. Would rather it is in Cloudron, but not sure I am competent enough to build it as custom app. Maybe will have a bash.

@peterle Yes, sure. We need to bring up the package at https://github.com/M-arcus/friendica-docker-image up to speed. It was packaged 3 years ago, so it probably won't work now.

@why42 yes I think this would not be suiteable for LAMP. But having a custom app package for custom scripts is very much doable if you have used docker at some point. Otherwise we are happy to help.

Here is my workaround for the Drupal 9.1.4 installation on the Cloudron LAMP app (Paket v1.4.0-1).

Install the LAMP app on your Cloudron Update the app ressources to minimal 1.5 GB RAM open the terminal switch to the www-data user su www-data change directory to /app/data cd /app/data install drupal via composer composer create-project drupal/recommended-project move everything from inside recommended-project into the root of /app/data mv recommended-project/* . & mv recommended-project/.* . remove the default public folder and make a symbolic link rm -rf public& ln -s web public open your browser and install drupal the last submit ends with an error in the terminal install drush via composer composer require drush/drush clear the cache php vendor/bin/drush cr open your drupal installation again. #yippie

Maybe there is an unknown error in the drupal installation process or something between mysql/drupal/cloudron. No hints in the logfiles. But this workflow works for me.

@girish Thank you

@nebulon I think for a start just making ssh based git push/pull work with hooks on the Github Pages and LAMP app is good enough. I think for rest of the "complicated" apps, we have to really make some container based PaaS, where we build the container as well (just like Heroku). That's a really long term feature.

@girish sorry for spamming this old thread. I added a complete new post: https://forum.cloudron.io/topic/4203/git-push-ci-cd-workflow-for-the-lamp-app

apt install -y cron apache2-dev php7.4 php7.4-{bcmath,bz2,cgi,cli,common,curl,dba,dev,enchant,fpm,gd,gmp,imap,interbase,intl,json,ldap,mbstring,mysql,odbc,opcache,pgsql,phpdbg,pspell,readline,soap,sqlite3,sybase,tidy,xml,xmlrpc,xsl,zip} libapache2-mod-php7.4 php-{apcu,date,geoip,gettext,imagick,gnupg,mailparse,pear,redis,twig,uuid,validate,zmq} ghostscript libgs-dev ffmpeg && \

is what we install. exiftool is there in the base image.

The latest LAMP app now has LDAP addon enabled. For existing installations, LDAP will still be off (sorry).

@girish Thank you girish

@girish said in PHP mailparse extension/modul:

@andirahmat OK, I learnt a bit more about this now.

PECL is PHP extensions written in C. PEAR is PHP extensions written in PHP.

Luckily, as you pointed out mailparse is already there in apt. So, I have pushed an update for the PHP 7.4 app to have mailparse in it. I also added it to the php-postgres repo.

Thank you! I'll try again.

@p44 The MySQL database in Cloudron is shared by all the apps. In MySQL 5.7.5+, they changed the behavior to have ONLY_FULL_GROUP_BY to be the default to be more compliant. Disabling this at the server level will cause issues for other apps.

The correct way to fix this is to have the app disable this mode at a session level. Like https://dev.mysql.com/doc/refman/5.7/en/group-by-handling.html . Essentially, the app has to call "SET GLOBAL sql_mode=(SELECT REPLACE(@@sql_mode,'ONLY_FULL_GROUP_BY',''));" after creating a connection.

So, I guess the answer is:

Either sendy has to provide a mechanism to do the above. Alternately, you cannot run sendy in Cloudron. You have to create a dedicated server for this and setup MySQL accordingly.

@girish Hmm indeed, that's bad... couldn't some kind of basic auth be used instead a bit like the Youtube-DL app ?

@girish you're a genius man! I think it should work, I'll try it and come back. Thanks!!

This is unlikely to be supported by a specific LAMP app. In such a case it is probably better to build a customized app package following the documentation at https://cloudron.io/documentation/custom-apps/tutorial/

Currently the Lamp app still does not have LDAP enabled for the same reasons about the username:password mentioned in the linked forum thread.

@girish & @jdelon02

I believe LAMP app supports cron, in that case use the following script to auto-pull from a git repo whenever an update is pushed to that repo:

Put in /app/data/updateGit

#!/bin/sh ## CD to the path where the app is installed. cd /app/data/public UPSTREAM=${1:-'@{u}'} LOCAL=$(git rev-parse @) REMOTE=$(git rev-parse "$UPSTREAM") BASE=$(git merge-base @ "$UPSTREAM") if [ $LOCAL = $REMOTE ]; then echo "Up-to-date" elif [ $LOCAL = $BASE ]; then echo "Need to pull" git pull elif [ $REMOTE = $BASE ]; then echo "Need to push? Some files changed / commited here, this should not have happened." else echo "Diverged?? Help!" fi

Then in /app/data/crontab:

*/5 * * * * bash /app/data/updateGit 2>&1 ## Or if you wanna avoid filling logs: */5 * * * * bash /app/data/updateGit 2>&1 > /dev/null

This will check if a pull is needed from a repo every 5 minutes. (Re: https://crontab.guru/every-5-minutes)

You can also modify the bash script to rebuild, npm install, composer install, etc when needed too.