> For the complete documentation index, see [llms.txt](https://docs.edgegap.com/llms.txt). Markdown versions of documentation pages are available by appending `.md` to page URLs; this page is available as [Markdown](https://docs.edgegap.com/~/changes/305/learn/orchestration/private-fleets.md).
# Private Fleets
{% hint style="success" %}
[**Registration for Private Fleets and Server Browser is open**, book your seat today!](https://forms.gle/LGtHP46JPaPhLg787) Limited availability.
{% endhint %}
Reserve standby capacity on monthly basis for persistent 24/7 always online servers. Choose your preference between Virtual Machines (Performance) or Bare Metal (Overdrive) specifications.
**Private Fleets are an optional addition to Edgegap Cloud (default),** a popular hosting option enabling [Persistence](/~/changes/305/learn/orchestration/persistence.md) with 24/7 always online servers and favorable pricing for games with predictable traffic.
🌟 [**Upgrade to Pay as You Go tier**](https://app.edgegap.com/user-settings?tab=memberships) **to unlock private fleets hosting with standby capacity!**
{% hint style="success" %}
Alternatively, see [Deployments](/~/changes/305/learn/orchestration/deployments.md#match-bound) orchestration to utilize fractional vCPU pricing.
{% endhint %}
{% hint style="warning" %}
[Smart Fleets (legacy)](/~/changes/305/docs/fleet.md) and [Sessions (legacy)](/~/changes/305/docs/session.md) will be deprecated and replaced by [Private Fleets](/~/changes/305/learn/orchestration/private-fleets.md) and [Server Browser](/~/changes/305/learn/server-browser.md) on March 1st 2026. [**Reach out over Discord**](https://discord.gg/MmJf8fWjnt) **to learn more and get help with migration.**
{% endhint %}
## ✔️ Introduction
Private Fleets let you pick specific host locations and reserve dedicated standby compute capacity within our globally available cloud infrastructure, suitable for:
* [**persistent 24/7 servers**](/~/changes/305/learn/orchestration/persistence.md) featuring open worlds with user-generated content placed by players,
* **MMO games** prioritizing social experiences with high amount of concurrent users per server,
* **released games with mostly predictable daily traffic** patterns and stable player communities.
## ☁️ Hardware
Pick one of our generic host specifications:
| Specification | Performance | Overdrive |
| ------------------------------------------------------------------------ | ----------------------------------------------------------------------------------------------------------------------------------------------: | ----------------------------------------------------------------------------------------------------------------------------------------------: |
| Best Suited For |
match-based games, or with predictable traffic
|
CPU-intensive games (AI), or above 1:2 CPU/RAM ratio
|
| RAM Specifications | 32 GB | 64 GB |
| Free Egress Included | 5 TB per Host | 5 TB per Host |
|
Price, pre-paid monthly
(commitment based)
| **$250.00 per Host** | **$290.00 per Host** |
[Preview a pricing estimate during configuration of your fleet](https://app.edgegap.com/private-fleet-management/private-fleets/list). Your first month reservation price and free egress quota are pro-rated. Fleet reservation is automatically renewed on the first day of each month for all active hosts in your fleets.
You may designate any number of reserved private hosts for deletion prior to the renewal date. Deleted hosts will not be charged for the subsequent renewals.
In the event of delayed host start or insufficient provider capacity, your invoice for the current and renewal periods will be adjusted accordingly. Network egress exceeding included capacity of 5 TB per private host is billed at public cloud egress price.
{% hint style="success" %}
A fleet with [active hosts](#lifecycle) may not swap specifications. Create a new fleet or delete active hosts first.
{% endhint %}
{% hint style="info" %}
Small amount of resources on each host is reserved for Operating System, Monitoring, and Orchestration.
{% endhint %}
## 📍 Locations
Choose a centroid by dropping a pin and preview your primary/fallback locations for new hosts.
Select a fallback strategy to control suitable locations for each centroid:
* **Proximity (default)** - 700 km centroid radius, for more latency-sensitive designs,
* **Redundancy** - 1,400 km centroid radius, allowing further locations as fallbacks.
{% hint style="info" %}
Location availability depends on chosen hardware specification. Overdrive availability is limited.
{% endhint %}
📢 API provisioning under development, [let us know](https://discord.gg/NgCnkHbsGp) about your Continuous Integration needs!
## 🔄 Host Lifecycle
Private Fleet hosts follow a simple three step lifecycle sequence.
📢 Scheduling new host additions is under development, [let us know](https://discord.gg/NgCnkHbsGp) about your release process!
#### 1. Pending
New hosts have been requested, we're starting them in order of distance from centroid.
#### 2. Active
Host is active and responsive, ready to [/pages/0UXQAhtFuL0FkdoUmYTh#id-1.-start-a-deployment](https://docs.edgegap.com/~/changes/305/learn/orchestration/pages/0UXQAhtFuL0FkdoUmYTh#id-1.-start-a-deployment "mention").
#### 3. Scheduled for Deletion
While stile active, this host will be deleted at the end of the current billing period (month). This action can be reverted (cancelled) at any time until deletion takes place.
{% hint style="warning" %}
Deployments running on private hosts will be [gracefully stopped](https://docs.edgegap.com/~/changes/305/learn/orchestration/pages/0UXQAhtFuL0FkdoUmYTh#id-5.-deployment-stopped) once the host is deleted.
{% endhint %}
{% hint style="success" %}
**Prevent frustration due to rollbacks and gameplay disruption** by notifying players and rerouting new deployments to other private fleet hosts, once a private host scheduled for deletion is nearing shutdown.
{% endhint %}
## ⚡ Deployment
[Deployments for Private Fleet](/~/changes/305/docs/api/dedicated-servers.md#post-private-fleets-deploy) work exactly the same as [regular cloud deployments](/~/changes/305/docs/api/dedicated-servers.md#post-deployments), except you must specify a prioritized [list of host IDs](/~/changes/305/docs/api/dedicated-servers.md#get-private-fleets-fleet-name-hosts) to deploy to.
* Deployment will be started on the first private host in your list with sufficient resources.
* If no private host has available capacity, deployment will [#overflow-to-cloud](#overflow-to-cloud "mention").
### Overflow to Cloud
If your selected private fleet hosts don't have sufficient capacity at the time of deployment, your request will automatically use [Deployments](/~/changes/305/learn/orchestration/deployments.md#server-score) strategy and deploy to Edgegap Cloud.
To account for this possibility and prepare:
* Read [Deployments](/~/changes/305/learn/orchestration/deployments.md#resource-specifications) injected variables to identify overflow case within deployment.
* Cloud hosts may provide lower CPU frequency than private host Overdrive [#hardware](#hardware "mention"), restrict player capacity or other CPU-intensive features in server to prevent bottlenecking.
* Substitute missing private host beacon with one of the low latency cloud [Ping Beacons](/~/changes/305/learn/orchestration/ping-beacons.md).
* Prepare for your deployment [gracefully shutting down](https://docs.edgegap.com/~/changes/305/learn/orchestration/pages/0UXQAhtFuL0FkdoUmYTh#id-5.-deployment-stopped) after 24 hours (max cloud lifetime).
* Manage [Persistence](/~/changes/305/learn/orchestration/persistence.md) and [Persistence](/~/changes/305/learn/orchestration/persistence.md#recovery-objectives) to prevent data loss.
## 🔎 Discoverability
Find deployments running on private fleets and establish player sessions using [Server Browser](/~/changes/305/learn/server-browser.md), or [Deployments](/~/changes/305/learn/orchestration/deployments.md#filter-deployments) to integrate a custom backend / third party solutions.
📢 [Matchmaking](/~/changes/305/learn/matchmaking.md) integration under development, [let us know](https://discord.gg/NgCnkHbsGp) about your needs!
### Ping Beacons
Enable beacons for your fleet to automatically deploy and maintain highly-available [#ping-beacons](#ping-beacons "mention") on each of your private hosts. Hosting beacons on the same machine as your deployments will provide the most precise ping measurement.
[^1]: base frequency
[^2]: turbo - short CPU bursts
