Allow tweaking via a single user controllable file to customize some instances that the default settings may not provide for.

This touches a number of items and is the basis to allow them to be possible, including:

• #25 - allow customization of php.ini max sizes
• #23 - support multiple php versions
• #11 - allow customizable defaults for wp-config
• #9 - add ability to add custom php packages
• #3 - allow custom default user/pass for wp

http://wp-cli.org/config/

Any way to have it respect the host's preferences here?

i.e. an elasticsearch or redis container that we can interact with be amazing

Currently the docs say this:

Add the bin path in the repo to your OS's include path.

The documentation should be expanded with some more info on how to do that.

I believe this is a side effect from running in a container:

Warning message is displayed before the output of some WP-CLI commands, like wp plugin list

Having define( 'JETPACK_STAGING_MODE', true ); in the wp-config.php file may resolve issues with Jetpack on production sites becoming disconnected after importing a production database to a local site.

Currently working on a project that requires a large overhead of memory available. Need to be able to set this pretty high (higher than we'd want for a default, I'm sure)

currently just gave me a 500 error in browser

Just a possibility, but in using both, I've been finding Redis to be much more capable and easy to deal with.

http://stackoverflow.com/questions/10558465/memcached-vs-redis

http://www.infoworld.com/article/3063161/application-development/why-redis-beats-memcached-for-caching.html

Change value in php.ini in config folder, it doesn't show after a full dockerbox restart in a phpinfo() call

32m might be appropriate for prod for most sites, but may be a bit low for what we're doing on dev.

I would prefer a higher default or a way to change without forking wpdocker - a custom settings file seems to be a recurring necessity here to be able to customize these types of things

Memcached is not fully compatible with PHP 7 yet, although it works fine for some sites. However, on one of my local sites, enabling object caching with the PECL Memcached extension brought the site to a crawl. The old Memcache extension is less desirable for object caching, but a PHP7-compatible port exists so that's what I'll be using by default instead of Memcached. The Memcached extension will still be available and can be used on an opt-in basis.

Make mariadb/mysql use utf8mb4 as their default/typical encoding:

When creating a new site, the site name should be used in the WP_CACHE_KEY_SALT constant. Otherwise cache collisions may occur. Currently it is set to sitename01 for all new sites.

i.e. for non-WP type sites (or if rebuilding a broken one from a docker rebuild), some commands to create a database for example would be helpful

Putting this here in case it's a setup issue - I'm importing a rather large database at around 4.0 Gb and got this error that I'm looking into. Perhaps it's a system level issue, but figured I'd point this out in case anyone else has run into this problem:

2017-02-25 23:07:21 140083114277632 [ERROR] InnoDB: The total blob data length (10709816) is greater than 10% of the total redo log size (100663296). Please increase total redo log size.

After logging into mysql and running the following, it looks like we have approximately 268 Mb allocated here.

pilothouse compose exec mysql /usr/bin/mysql -u root -proot
Welcome to the MariaDB monitor.  Commands end with ; or \g.
Your MariaDB connection id is 7
Server version: 10.1.21-MariaDB-1~jessie mariadb.org binary distribution

Copyright (c) 2000, 2016, Oracle, MariaDB Corporation Ab and others.

Type 'help;' or '\h' for help. Type '\c' to clear the current input statement.

MariaDB [(none)]> SELECT @@innodb_buffer_pool_size;
+---------------------------+
| @@innodb_buffer_pool_size |
+---------------------------+
|                 268435456 |
+---------------------------+
1 row in set (0.00 sec)

MariaDB [(none)]> SELECT @@innodb_buffer_pool_instances;
+--------------------------------+
| @@innodb_buffer_pool_instances |
+--------------------------------+
|                              8 |
+--------------------------------+
1 row in set (0.00 sec)

MariaDB [(none)]> SELECT @@innodb_buffer_pool_chunk_size;
ERROR 1193 (HY000): Unknown system variable 'innodb_buffer_pool_chunk_size'

The innodb_buffer_pool_chunk_size not existing seems like it could be an issue too, but I'm not sure. I'll post more if I figure out a solution here.

On my first run of dockerbox up, I get stuck with the error No container found for php_1 repeating for several minutes until I finally cancel the run. Just before that is a problem with mailcatcher, but I don't know if that's the cause or just a red herring.

Here's the messages just before and after the error starts:

ERROR:  Could not find a valid gem 'mailcatcher' (>= 0), here is why:
          Unable to download data from https://rubygems.org/ - SSL_connect returned=1 errno=0 state=error: certificate verify failed (https://api.rubygems.org/specs.4.8.gz)
ERROR: Service 'mailcatcher' failed to build: The command '/bin/sh -c apk add --no-cache --virtual .build-deps     build-base curl ruby-dev && apk add --no-cache curl libstdc++ ruby ruby-json sqlite-dev && gem install mailcatcher --no-ri --no-rdoc && apk del .build-deps' returned a non-zero code: 2
ERROR: No container found for php_1
ERROR: No container found for php_1
Waiting for MySQL to become ready... ERROR: No container found for php_1
ERROR: No container found for php_1

If the command is run from within a site directory, the generate-ssl command should detect that and not prompt the user.

We should be able to view error logs without having to open a shell in the host. Add a log directory in the dockerbox root where logs can be synced from the containers.

Not sure if you want to do this, but I find myself doing this every time to set up a pre-existing project

This may be a misunderstanding of use on my part: I'm unable to get breakpoints to trigger for XDebug when the code I'm running is executed on a CRON.

I figured this is due to the XDebug setup using cookies that do not exist for the CRON when it's running in the background. Is there a setup that can be implemented to enable this, similar to how the wp-cli command needs a flag; something that can be set at the system level?

Support multiple versions of PHP, switchable via a cookie like Xdebug is currently toggled.

MySQL database data should be persistent even if the container is deleted. Data Volumes (https://docs.docker.com/engine/tutorials/dockervolumes/#/data-volumes) provide this functionality. We should store the mysql data directory in a persistent volume.

SoapClient is used by some WooCommerce modules; should be available by default.

This should be added to the nginx shared.conf.inc so that it doesn't pass along requests to static files to php (which prevents additional pointless processing for php, etc):

location ~* \.(?:ico|css|js|gif|jpe?g|png|svg|woff|ttf|eot)$ {
    expires 6h;
    add_header Pragma public;
    add_header Cache-Control "public";
}```

Existing VVV nginx config tries php before .html:

similar to how we can run wp in the host, add support for codeception

I need to change the name of the project. http://wpdocker.com/ is already an established commercial entity, and I don't want to cause confusion with that, or to be using a name they may have trademarked.

When I originally named the project, I did a Google search for wpdocker and didn't see any relevant results, so I assumed no one was using the name. What I didn't notice was that Google oh-so-helpfully changed my search query to "wplocker", thus I didn't see wpdocker.com in the search results.

So, I've settled on "DockerBox" as the new name of the project. This more accurately reflects the goal of the project to be a more generic local development environment using Docker, useful not only for WordPress projects, but general PHP projects as well.

As part of this ticket, I will be updating all instances of wpdocker in the codebase to dockerbox. I own the DockerBox GitHub organization, and will be moving this repository to https://github.com/dockerbox/dockerbox shortly. GitHub will redirect the old repository URLs, and all Git history, issues, etc. will be retained. Users will still be able to clone, fetch, and push to https://github.com/philipnewcomer/wpdocker as GitHub will redirect those requests to the new location (https://help.github.com/articles/renaming-a-repository/).

It would be nice to be able to enable xdebug processing with a simple ?xdebug=on flag to the request URL.

If the site is giving a 500 error then there's no option to use set a cookie yet and it would be nice to use xdebug to track down what the issue is.

For example, I prefer and have always used dev/dev as it's fast and easy to type with one hand.

I'd love to be able to have that be my default - maybe a local file or .env file?

Create a Chrome extension (and possibly for other browsers as well) to show and toggle the current status of the Xdebug cookie

I know this worked before, but it's no longer working:

Other WP commands aren't working either:

This is with the latest version of DockerBox

The MySQL, Nginx, PHP, and Memcached containers should all be able to be restarted individually using the wpdocker command.

Subject Alternate Name is now a required field for Chrome to trust a certificate, and is not currently included in the SSL certificates generated by PH. This results in local sites using SSL not being accessible in Chrome 58.

Details:

• https://www.reddit.com/r/sysadmin/comments/66clon/chrome_58_displays_an_untrusted_warning_for_certs/
• webpack/webpack-dev-server#854
• https://deliciousbrains.com/https-locally-without-browser-privacy-errors/

I have no other machines running but I'm getting this error:

ERROR: for nginx Cannot start service nginx: driver failed programming external connectivity on endpoint pilothouse_nginx_1 (b4652d1471ea2aa201116ffcb3d17d3ad26456541b86838a2350be64c89bd2a8): Error starting userland proxy: Bind for 0.0.0.0:80: unexpected error (Failure EADDRINUSE)

Any ideas?

Composer should be run inside a PHP container, instead of on the host. This will make the entire stack more portable, and eliminate issues with a package requiring a PHP version higher than what is currently installed on the host.

Will also help with cross-platform compatibility by not requiring PHP to be installed on Windows and Linux hosts where it is not available by default.

WP-CLI should be actually installed and managed in/by the container instead of being copied from the host where it is managed via Composer.

looking for advanced cache by default causes some php warnings:

Warning: include(/var/www/html/site/htdocs/wp-content/advanced-cache.php): failed to open stream: No such file or directory in /var/www/html/site/htdocs/wp-settings.php on line 86

Warning: include(): Failed opening '/var/www/html/site/htdocs/wp-content/advanced-cache.php' for inclusion (include_path='.:/usr/local/lib/php') in /var/www/html/site/htdocs/wp-settings.php on line 86

The fix for #5 had the side effect of removing formatting (colors, tables, etc.) for wp-cli command output, due to the removal of docker's pseudo terminal emulation, or something like that. We need to get that formatting back without reverting #5.

readme says admin/12345 but it's actually admin/password

To avoid potential conflicts, please ensure there's an empty line prior to adding to the hosts file, and that each entry gets its own line.

When my first site was added to /etc/hosts, it was appended to the last line in the file. Since another program edited the hosts file and the line ended with ##, it commented out the new host entry that dockerbox created.

## Local by Flywheel - End ##127.0.0.1 mdub.dev #dockerbox

should be

## Local by Flywheel - End ##

127.0.0.1 mdub.dev #dockerbox

For example, I'd like all my installs to have display errors set to true by default: