It seems to be a permission issue regarding the install.log file situated in public/legacy/ :
2021-09-13 09:19:46...ERROR: Cannot create writable dir cache/upload/import
2021-09-13 09:19:46...installerHook: Info: custom/install/install_hooks.php not present, no custom hooks to execute
I really don’t know what else could it be, but also don’t know what other permissions could I set instead of [myusername]:staff as it is the ones set on the root folder and into MAMP Pro ?
Thank you for your time, we can wait to start using SuiteCRM 8 !
From first glance it seems that suite is not able to write to those folders.
Are there any logs on public/legacy/suitecrm.log that seem relevant?
One thing to have in mind is that the Beta 3 zip (unlike the Beta 2) has an inner folder inside the zip, where the beta 2 had the project files directly on the root level of the zip.
How does your structure look like from the root? is SuiteCRM 8 (core, cache, etc) directly on the root folder?
I tried to set 127.0.0.1 as the database host name and the only thing that changed is that I can see a Symfony debugger bar at the bottom of the project now :
Thu Sep 16 08:28:35 2021 [65586][-none-][FATAL] Could not connect to DB server localhost as root. port 3306: No such file or directory
Database still not created…
It looks like I’m not using the correct mysql as I’m running SuiteCRM via Mamp Pro.
Installation script ./bin/console suitecrm:app:install may be using the mysql installed by hand via the command line and not the one used by Mamp Pro ?
I’m currently looking for a way to find out which one is used (mysql).
Thank you very much for your help, I hope to find out what I’m missing
Thank you for the resources you gave me, it helped a bit.
Just succeed the install using the “pre-built” package, I don’t know WHY it now works, but I think it could be because I deleted the mysql I installed using brew before, still using 127.0.0.1 instead of localhost.
Here’s the history of the CLI install :
SuiteCRM Silent Install
============
Please enter the db username: root
Please enter the db password:
Please enter the db host: 127.0.0.1
Please enter the db name: suite8
Please enter the admin username: max
Please enter the admin password: ***
Please enter the suite 8 address: http://dev.suite8/
Install demo data:
[0] yes
[1] no
> 1
Step 1: Config Creation Complete
Starting SuiteCRM CLi Installation
SuiteCRM CLi Install Complete
Step 2: Legacy Installation Complete
I will update this post when I’ll figure out how I succeed in the install, hoping it’ll help some people in the same situation.
Just figured out that it may have worked because I changed localhost to 127.0.0.1 (or whatever associated IP address) following @clemente.raposo’s advice.
OMG… I completely forgot to mention that we use Git Plesk’s extension to fetch code from our repo on Github ! What a shame to forget something that important.
When I try to install SuiteCRM after a simple drag & drop to the server, it obviously works following the installation guide.
However, when I use ps -ef it’s www-data that indeed runs the server.
EDIT : Just reinstalled the Plesk image and trying to find out what’s going on with the Git fetch… Maybe I’m just ignoring some files that I shouldn’t ignore.
EDIT 2 to clarify the context: We initialized a .git locally and linked it onto a remote. We pushed the SuiteCRM project before the installation and fetched it on the distant server from the remote. After the fetch on the distant server, I then add by hand the default .env and extension/ folder before trying to install via SSH
I’m now ignoring nothing but the files already ignored by the .gitignore located in the repo and then committing everything to the Git before fetching from the Plesk server.
I reinstalled the same Plesk image, installed the needed php extensions and followed the developer installation guide :
Permissions still the same for the cache/ folder :