angry-dan / vdd Goto Github PK

add memcache settings to standard settings - but only if memcache module is present in sites/all/modules/contrib

Current VDD uses a 32 bit Ubuntu precise as a base - 12.04 which Acquia just migrated away from

We need a 64 bit Ubuntu Trusty base instead - 14.04

Have CSS & JS compile automatcially - needs either settings per site or a careful best practice standard for folder locations.

Currently we only get one solr environment, solr4 - we should provide solr 3 as well as solr4 since the likes of Acquia still use solr 3

Everyone should have mongodb locally!

Work out how to use SSL so we can have HTTPS connections without errors

needs memcache

Current solr release is 4.10.3 but if I specify that then solr cores don't work. This is probably due to missing config files

varnish needs to be started after vagrant up at the moment using

vagrant ssh
sudo /etc/init.d/varnish start

Create simple instructions for setting up Drupal through syslog and how to see the results in pimpmylogs.

Hopefully this config can be done automatically during vagrant provision and necessary conf added to the sites settings.php

They are banned by Drupal coding standards and masked a bug that Adam and Steve were having the other day.

phpmyadmin stopped working after mariadb update

Add varnish to port 8080

needs a hero's vimrc file, like mine

sql-cli doesn't work for reasons that I don't fully understand, but I wish it did! Currently you have to ssh in then you can drush @site sql-cli

Pantheon uses redis rather than memcache - this should be available on vdd as well

Its got things in it per site that are never used...

"account_name": "root",
"account_pass": "root",
"account_mail": "[email protected]",
"site_mail": "[email protected]",

To go a bit further, I wondered if perhaps we should scrap the config.json route of adding sites to VDD at all. Since we have defined the directory projects are stored in (default is ~/Sites) we could just recurse over the folders in that dir and look for ones called [shortname].dev, making a site for each. Specific config could reside in an optional .vdd/config.json file in the project root meaning checking out the project code into ~/Sites and a reprovision would be enough to add it.

Create a centralised settings file per site like Acquia do. This would be like the aliases file but could inlcude things such as:

• Base URL
• Apachesolr configuration defaults

running drush master-execute causes all aliases environments to be contacted by drush

As the vagrant box has no ssh keys for the other environments e.g. production it can't do anything to it but it generates errors.

drush uses @self when executing child processes - it must be this causing issues.

apache crud and old and has rubbish error reporting.

nginx is new, sleek and has error reporting.

can we swap completely? too controversial?

MailCatcher needs so settings in php.ini provided by the vdd_php.ini file which is presently only being included in apache's php config. This needs including in php-fpm as well

What would make VDD really awesome is some kind of vdd cli...
(Symfony based of course)

So:

$cd ~/Sites/some-project
$vdd install
Project shortname? [some-project]
Path to docroot? [docroot]
Import a DB? #Paste a DB backup path here. .gz files are decompressed automatically.
Re-provision machine now? [yes]
Done! Visit your new site: http://some-project.dev

And off it goes... backup and update config.json then re-run vagrant --provision.

How cool would that be?! Steps to set up a site go down to just:

git clone
vdd install, enter, enter, cmd+v, enter
click link

I've edited the drush.rb file to checkout drush 6.2.0 and make that the default drush command

Also needs a drush7 command on its own codebase for running bleeding edge latest version for Drupal 8.