A private cloud built on dedicated hardware, hosted in our Carthage
datacenter or inside your own walls, and operated either by the FelCloud
team or directly by yours.
Dedicated hardwareData under local law100% open sourceYou choose who operates it
The platform is the same in all three. What changes is where the machines
sit and whose hands are on them day to day.
Most chosen
Hosted Private Cloud
In our Carthage datacenter · Operated by FelCloud
Your own hardware in our datacenter, isolated from every other
customer, with our team on call. The fastest way to a private cloud:
no racks to buy, no datacenter to run.
Bare-metal nodes reserved for you, no shared tenancy
Private region with its own network and its own storage
Operated, patched and monitored around the clock by our team
Invoiced in dinar, euro or dollar, with a contractual SLA
The machines never leave your building, and you still do not have to
run them. We install the platform in your racks and operate it
remotely, with on-site interventions when hardware needs hands.
Your racks, your power, your physical access control
Design, installation and commissioning by our engineers
Day-to-day operations, upgrades and on-call handled by us
Remote access is nominal, logged, and revocable by you at any
time
Not a label on a brochure. Four things you can check in the contract.
Data stays where you put it
Carthage, Tunisia, or your own premises. No replication to a region
you did not ask for, and the location is written into the contract.
One jurisdiction, one law
A Tunisian operator under Tunisian law. No foreign parent company
that an extraterritorial order could compel to hand over your data.
Open source, end to end
OpenStack, Ceph, Kubernetes, Linux. No proprietary layer that only
we can operate, and no licence that expires out from under you.
Reversible by design
Standard APIs and standard image formats, so leaving is an export,
not a rewrite. If the contract ends, the platform keeps running.
The same platform we run
Your private cloud runs the stack behind our public cloud, so the
console, the API and the automation you already know work unchanged.
OpenStackCompute, network and block storage, driven by the same API as our
public region.
Ceph storageReplicated block and object storage, with no single machine holding
the only copy.
KubernetesManaged clusters on request, on top of your own private
region.
Terraform & PulumiInfrastructure as code from the first day, not bolted on
afterwards.
Backup & recoveryScheduled snapshots, off-site copies and a disaster-recovery plan
that gets tested.
Monitoring & logsMetrics, alerting and retained logs, readable by your team as well as
ours.
Side by side
The questions customers actually ask before signing.
Criterion
Hosted Private Cloud
On-Premises, Managed
On-Premises, Self-Operated
Where the machines sit
Our Carthage datacenter
Your datacenter
Your datacenter
Hardware bought by
FelCloud, rented to you
You, or financed with us
You
Day-to-day operations
FelCloud
FelCloud, remotely
Your team
Who holds administrator access
Both, each with their own named accounts
Both, ours nominal and revocable
You only
Upgrades and patching
Scheduled with you, carried out by us
Scheduled with you, carried out by us
Your team, with our releases and runbooks
Runs disconnected
No
Partly, operations need a link
Yes, fully air-gapped
Typical lead time
2 to 4 weeks
4 to 8 weeks, hardware depending
6 to 10 weeks, training included
Best fit
Teams who want isolation without running a datacenter
Regulated bodies required to keep data on their own site
Organisations with their own operations team and no outside access
How a project runs
No private cloud is bought from a price list. Here is the path, and what
each stage actually costs you in time.
1
Scoping workshop
We go through your workloads, your compliance obligations and your
growth over three years. Free, and you keep the write-up whatever you
decide next.
Half a day
2
Sizing and quote
A written architecture, the bill of materials, the operating model and
a fixed price. One document, no hidden line items.
About one week
3
Build and acceptance
We deploy, we test with you against the acceptance criteria we agreed,
and we migrate a first real workload together.
2 to 8 weeks
4
Operate, or hand over
Either our team takes over operations, or we train yours and transfer
every credential, runbook and procedure.
Ongoing
Who asks us for this
Organisations whose data cannot leave the country, or cannot leave the
building.
Banks & fintechHospitals & health dataPublic sector & ministriesIndustry & manufacturingTelecom operatorsUniversities & research
Straight answers
How is this different from your public cloud?
The public cloud shares physical machines between customers and is
billed by the hour. Here the hardware is yours alone, the capacity is
fixed and reserved, and the contract is annual. Same API, same
console.
Can we start on the public cloud and move later?
Yes, and most do. Because both run the same platform, moving a
workload across is an image and volume transfer, not a rewrite of your
automation.
Who can log in to our machines?
Whoever you decide. Our engineers hold named accounts, every session
is logged, and in the self-operated model we hold none at all. There
are no shared or service accounts belonging to FelCloud.
What happens if we end the contract?
On your premises, the platform keeps running and we hand over every
credential and procedure. In our datacenter, we export your images,
volumes and object data in standard formats, then wipe the hardware to
a written procedure.
Why is there no price on this page?
Because a number here would be fiction. The price follows the number
of nodes, the storage, the network, the operating model and the
service level. The scoping workshop gives you a real figure, in
writing, and it costs you nothing.
Tell us what cannot leave the country.
Send us a line about your workloads and your constraints, and we will
come back with an architecture and a price. Or pick up the phone and
talk to the engineers who would build it.