Choose
your layer.
On the outside you keep control, on the inside we take over everything — from the full Proxmox cluster to the database cluster. Cheaper than your public cloud, always dedicated.
Contains layers 01–02. On this layer HOST SMITH takes over your network perimeter: a managed OPNsense firewall cluster in front of your VMs and clusters. OPNsense is the open-source firewall for rules, NAT, VPN and intrusion detection — with us always redundant: two instances in a cluster, connected via CARP IPs. If one firewall fails, the second takes over in a split second, without connections dropping.
Dedicated resources, predictable costs — no egress surprises. Traffic flat rate or on demand.
By default no service provider outside Germany — and even here exactly one: the one physically maintaining the racks.
Redundancy, snapshots, multi-zone — built in, no add-on.
As standard. 99.995% or 99.999% — with an additional SLA with penalties.
Managed services — without limits.
Everything dedicated, with practically no RAM or CPU limits. You use the service — we forge, operate and scale it.
Single node or replicated — incl. backups & point-in-time rollback.
Managed, monitored, backed up — single node or replicated.
In-memory cache & data structures, dedicated and managed.
HTTP cache in front of your application — response times in microseconds.
Message queue as node or cluster — reliably delivered.
S3-compatible object storage — dedicated, in Germany.
Prerender service with backing storage — SEO for your SPA.
Column-oriented analytics database — single node or cluster.
Search & analytics — from single node to cluster.
MariaDB-compatible active-active cluster for true high availability.
PostgreSQL-compatible active-active cluster across multiple zones.
Missing something? We'll forge it — tell us what you need.
Your storage. Your rules.
Shared and synchronous across all nodes — or strictly isolated per node. In every speed class.
CephFS
Mirrored and synchronized across all nodes — the foundation for live migration and true resilience. If a node fails, your data keeps running.
Isolated storage
Local to the node, with no network in between — for workloads where every microsecond counts, and for clear data separation.
Resilience — built in.
CARP IPs
Failover IPs that automatically move to the redundant instance on failure.
Firewall cluster
OPNsense in a cluster — if one firewall fails, the second takes over. Instantly.
Proxmox redundancy
HA cluster with live migration — VMs move before users notice anything.
Multi-zone & multi-region
Redundancy across zones and regions — standard or tailor-made.
Incremental snapshots
Space-saving, tightly meshed, automatic — your history is always there.
Point-in-time rollbacks
Back to any point in time — in minutes, not night shifts.
Your backups additionally at a second, physically separate location — encrypted, versioned, independent of the cluster. Available for every layer.
Our DPA lists exactly one service provider: the one physically maintaining the racks.
No hyperscalers, no sub-sub-processors, no surprises. Hosted on ISO 27001-certified servers — in Germany as the standard, on request (multiple regions) also at other locations worldwide.
The technology behind it — briefly explained.
Proxmox VE
The open-source virtualization platform your VMs run on. Cluster-capable, with live migration and high availability — the foundation of every layer.
OPNsense
Open-source firewall for your network: rules, VPN, intrusion detection. With us always as a redundant cluster with CARP IPs — open (you change along) or closed (we change for you).
CephFS
Distributed file system that mirrors your storage synchronously across all nodes. If hardware fails, your data simply stays available.
Docker Swarm & Kubernetes
Container orchestration: distributes your applications automatically across multiple nodes and restarts them on failure by itself.
Portainer
Web interface for your containers: deploy, scale, read logs — entirely without a command line. Perfect for teams without their own ops.
HAProxy & Caddy
Load balancer and web server at the network edge: distribute traffic across multiple instances, terminate TLS and obtain certificates automatically — with us with distributed ACME storage via Redis.
Already running with us.
From over $4,500 per month at a public cloud provider to under €800 — with better, dedicated resources. The team keeps deploying via Portainer as usual, we operate everything below: cluster, firewall, storage, backups. Plus real consulting instead of ticket roulette.
AI workloads on a shared Proxmox cluster, fully managed from layer 01 to 07: VMs, firewall cluster, containers and Portainer from one hand. Consultancy hosting with architecture advice — and a strong cost deal for the 3-node cluster.
Complete platform hosting across layers 01–05: Proxmox, firewall cluster, edge and container cluster. On top, managed services with open access — Elasticsearch for spot search, Redis, RabbitMQ and a YugabyteDB active-active cluster. The team develops, we keep the platform in the wind.
The complete environment from layer 01 to 08 — from the Proxmox cluster through firewall, edge and containers to the managed services behind them. Open access for the team, redundancy and backups built in: the app scales, the learning content stays available.
High-traffic WordPress, maximally performance-optimized: Varnish cache, tuned database and dedicated resources instead of shared hosting. The result: one of the best performance ratings in the industry — even on trade-fair weekends with load peaks.
Music licensing with high sales volume — fully managed: we are the DevOps department, from deployments through databases and caching to S3. The team keeps pushing code itself, we take over everything else. Every minute of downtime costs revenue — hence redundancy on every layer.
Distributed hosting of custom software on Windows servers — an edge case hardly anyone offers. 100% managed and fully automated: provisioning, updates and monitoring run without manual intervention.
SaaS platforms, agencies, entire company environments — many of our customers prefer not to be named publicly. We're happy to tell you in a conversation what we operate for companies like yours.
People instead of hold music.
HOST SMITH grew out of a digital product forge — we've been building platforms ourselves for years and know what good hosting means in everyday life. That makes us the consultancy hoster: we don't just operate, we advise, migrate and think along. With us you talk to people who truly understand your architecture. And if no standard layer fits, we develop yours together: turnkey, from idea to operation.
Get free advice
Developed and operated in Germany by Ambos Development — your infrastructure runs dedicated on ISO 27001-certified servers, in Germany as the standard, on request (multiple regions) also at other locations worldwide. Made in Germany.
Say hello.
Send us your email. We'll quickly figure out which layer fits you — or forge a new one.
