Official Introducing our new Hybrid Server Density workflow

Hi all, we've just released some workflow changes to Game Server Hosting, below is all of the information on what changes we've made and what they mean for you.

What is the new Hybrid Server Density workflow?
This new feature allows users to configure server density — the amount of servers that fit onto a machine — independently for both cloud and bare metal machines.

What are the main differences between this new feature and how server density was previously configured?
Previously, server density was controlled by setting "server usage settings" on a build configuration. Usage settings informed the scaling system of how much RAM and CPU should be allocated to each server. Server density was then determined by using the worst-case usage settings of all the build configurations within a fleet.

With the new workflow, server density is now explicitly configured at the fleet level by users.

Why have we made these changes / how does it benefit users?

To allow server density configuration to be more explicit and less confusing
The previous method of configuring usage settings on build configurations was confusing for a number of reasons:

1. It obfuscated the concept of server density. We had a page dedicated to it on our docs site but didn’t allow users to explicitly configure it, nor did we ever display or communicate in the UI what the server density on a fleets is configured to.
2. The ‘worst case scenario’ method caused great confusion, even to us internally! A fleet can have many active build configurations but only one of them actually controlled server density via their usage settings. The chosen configuration wasn't shown anywhere. This means a user would have had to potentially change usage settings on multiple different build configurations to affect their server density, but could never be sure which is actually taking effect or what exact effect it would have.
3. "Servers per machine” is a much clearer input to guesstimate when creating a fleet for the first time and don't yet understand the exact resource usage of your game server. Previously we asked users to input their build’s expected CPU and RAM usage when creating a configuration for the first time, information that likely isn't available straight away. However it's much easier to reason "I don't know resource usage yet, so I'll just run 1 server per machine initially", for example.

To facilitate Bare Metal server density configuration
The previous method of using build configurations and usage settings to configure server density, didn't differentiate between different types of machines. Bare Metal machines are hugely important for our enterprise customers, the new workflow allows server density to be properly configures across multiple machine types.


What will happen to my previously configured usage settings?
The new server density configuration settings have been auto-populated for all fleets. They have been calculated from the previous worst-case scenario build configuration assigned to each fleet.

This should mean there is no actual change to the server density of your fleets.

You can see the user guide for this feature in our docs here: https://docs.unity.com/ugs/en-us/manual/game-server-hosting/manual/guides/configure-server-density

If you do encounter any issues please reach out to our support team.

Thanks!

 

Hi Hamza, users do not have control over the cloud machine specifications available, this hasn't changed with the new workflow (we're just making that spec visible).

To control the amount of servers that will run per machine, please consult the user guide linked above

There are no changes to our pricing, which you can view here: https://docs.unity.com/ugs/en-us/manual/game-server-hosting/manual/concepts/pricing