Actually only the following files are served from nginx with proper HTTP headers for browser caching and character sets.

html|htm|jpg|jpeg|gif|png|ico|css|zip|tgz|gz|bz2|pdf|odt|txt|tar|bmp|rtf|js|swf|avi|mp4|mp3|ogg|flv|svg

Please a local font files with the following extensions too:

eot/ttf/woff/woff2

The t3a family of servers upgrade the old t2 family of servers. The t3a family is cheaper to run yet more powerful, so you'll save money and your clients will get an even better product.

Please enable it asap, as it will be beneficial for us serving the Middle East.

Thanks and regards

Hi,
Please allow to setup sphinx search on the server.

Would be handy to specify the timezone you would get in log files. For example apache logs are always in UTC and can be configured at the apache level.

Scaling up is easy.

Scaling down (with Vultr, DigitalOcean or Linode) requires to clone a server.

When cloning, server gets new IP address - this causes a lot of hassle with application integration or DNS configuration.

SUGGESTION:
It would be great to make scaling down hassle-free, so that after cloning server could have the same IP as previous server.

Version 1.20 provides some very useful switches such as error only logging etc.

Add the ability to turn on ssl_session_cache.

Gives a speed boost for but also cuts down overheads for not having to request new keys every time a page request is made.

This is a pretty standard option, it shouldn't cause any conflicts as can be set individually on drops.

This is one line for a quick gain...current SSL testing is failing on this point.

You support high cpu instances for AWS and Google Cloud but not DigitalOcean.

Can you include support for choosing different instance types for DigitalOcean such as their high CPU Instance?

Instead of forcing us to dig up the htaccess file for modifications, make this a simple Control Panel-based modification.

It would be nice to have a read only user for the database so that people that need to perform SELECT queries don't use the admin user to log in.

The server upgrade takes somewhere between 10-15 minutes and during the upgrade, visitors can't access the website. They see a 503 error.

I think Cloudways should show a maintenance page while the server upgrade is in the process.

Include a back-up estimate so we can tell what to expect on our bill. Have a bill est. that shows server cost(s) plus backups on each server.

There was an infrastructure update that defaults files/directories to 720 permissions. Now FTP users no longer see any of the files or directories because of these permissions. Chat support's "solution" is to use the "reset permissions" function, which will then change the permissions of files and directories to 770.

First, why were the default permissions changed from 664/755, to 720? I've never seen this configuration before, so it seems odd to me.

Second, why is this a two step process for FTP users to gain access to the public_html directory? The default application permissions should MATCH the "reset permissions" permissions.

A solution with rsync, sql parallelism and a load balancer will be great.
It is important for some sites to have 0% downtime.

I would like to have seperate settings per location in my nginx configuration. Currently I cannot access the "server.nginx" configuration file in my application (which I assume to be the site configuration).

I have GeoIP country switcher on my woocommerce site.
But when I use with Varnish, it forces back to the country which the GeoIP detects instead of the one user wahts to change to. Once Varnish deactivated, country switcher works properly. First detection correct as well as manual change of the country possible.
This module should solve the issue.
https://guides.wp-bullet.com/cache-woocommerce-currency-geoip-varnish-4-vmod/

Multi-server applications are very common these days.
Example: Application files on one server and Database on another server like RDS.
We need such option to put our large applications onto it.