Trellis Wordpress crashes with high CPU usage on Wordpress plugin install

trellis
1

Hi,

When I try to update Wordpress on Trellis, or install (any) plugin, Chrome browser window crashes, vagrant box is on high CPU for a few minutes, then abandons the request (upstream timeout).

I provisioned a new vagrant box as per the docs without any customizations except changing name and url of the box.

Trellis version: pulled at commit 11516d12e42355d0304eb4fb8c15be7ca8fcf6d4 today.
Host environment: Windows 7 x64 (latest updates)
Virtualbox: 5.0.14 (latest)

The site itself is a fresh unnmodified sage (an older version from ~2 months ago)

After clicking “update” in WP (in local Chrome browser), top shows php-fpm on high cpu for a few minutes. Local Chrome Window crashes, and the nginx error.log shows:

2016/02/04 09:08:30 [error] 27051#27051: *12 upstream timed out (110: Connection timed out) while reading upstream, client: 192.168.50.1, server: testproject.dev, request: "POST /wp/wp-admin/update-core.php?action=do-core-upgrade HTTP/1.1", upstream: "fastcgi://unix:/var/run/php-fpm-wordpress.sock:", host: "testproject.dev", referrer: "http://testproject.dev/wp/wp-admin/update-core.php"

/var/log/php7.0-fpm.log just shows no errors during that, and seems to be running fine.

Any idea? Thanks!

2

OK, so I found out this is simply a performance issue with the shared folders, like described in discussion “VirtualBox performance in windows”

The only way I can get decent performance is using vagrant with rsync instead of shared folders or NFS.

In Vagrantfile I replaced
# config.vm.synced_folder local_site_path(site), remote_site_path(name), owner: 'vagrant', group: 'www-data', mount_options: ['dmode=776', 'fmode=775']

with
config.vm.synced_folder local_site_path(site), remote_site_path(name), type: 'rsync', owner: 'vagrant', group: 'www-data'

There’s a bug in vagrant rsync though, find the fix here: [github.com]/mitchellh/vagrant/issues/6702