Self-managed runtimes / Self-operated bare-metal sandbox fleet alternative

Self-hosted bare-metal sandbox fleet alternative for Linux agents

A self-operated bare-metal sandbox fleet is a deployment model, not one vendor’s managed agent API. Physical disks can already retain an entire installed Linux system. The comparison is who operates isolation, scheduling, durable state, recovery and file access across many agent identities.

Start buildingCompare capabilities ↓

Reviewed 2026-10-19 · Published by the Sandbox0 team

When Self-operated bare-metal sandbox fleet fits

Choose a self-operated fleet when hardware ownership, custom kernels, device access or infrastructure control justify running the platform. An existing competent operations team and well-utilized capacity can change the economics substantially.

When to consider Sandbox0

Consider managed Sandbox0 for compatible Linux agents when durable whole-RootFS state, automatic online growth and supported paused files should be provided by the runtime. Sandbox0 Self Host is unsupported; BYOC is planned and unavailable, so this is not a guide to deploying Sandbox0 on your own servers.

Capabilities

Self-operated bare-metal sandbox fleet and Sandbox0: what persists?

Feature comparison with documentation and verification status
Environment and runtime lifetimeSelf-operated bare-metal sandbox fleetNot established in this reviewSandbox0ttl and hard_ttl default to 0 (disabled). Configure idle pause and deletion policies for the same persistent sandbox identity; an idle running sandbox still accrues compute charges.Read documentation →
Persistent filesSelf-operated bare-metal sandbox fleetThe physical-host reference includes retained local disks; fleet retention/recovery depends on the operator’s design.View evidence ↗Sandbox0The writable RootFS survives successful checkpoints, including software, system configuration, dependencies, and user data.Read documentation →
Whole RootFSSelf-operated bare-metal sandbox fleetLocal system disks can preserve the entire installed OS and writable paths.View evidence ↗Sandbox0Persistence extends beyond /workspace to paths such as /usr/local, /etc, and /home. Runtime tmpfs and virtual filesystems are excluded.Read documentation →
Release compute, keep stateSelf-operated bare-metal sandbox fleetA generic hardware fleet does not define one managed compute-release or process-resume contract.View evidence ↗Sandbox0Pause compute between tasks; resume on available capacity. Default resume starts fresh processes from the saved filesystem.Read documentation →
Memory and process stateSelf-operated bare-metal sandbox fleetMemory checkpoint support depends on the operator’s hypervisor/runtime, not on owning persistent disks.View evidence ↗Sandbox0Experimental memory checkpoints can preserve compatible process state. Use memory: true on both pause and resume on a compatible deployment. A filesystem checkpoint alone does not preserve running processes.Read documentation →

Online RootFS growth

On growth-enabled images, Sandbox0 automatically expands the writable RootFS as free space runs low, within a configured sparse logical ceiling up to 64 TiB. Agent processes keep running. Growth is enabled in the qualified Cloud region; confirm activation for other regions. Existing images and snapshots retain their captured storage geometry. The ceiling is logical capacity, not preallocated host disk.

An operator can design storage expansion, but local hardware capacity and filesystem management need their own automation and failure handling. The hardware floor alone does not provide Sandbox0’s integrated low-space-triggered online durable RootFS growth.

Files while compute is paused

Sandbox0's Files API reads persistent files without resuming compute. Filesystem-only pauses also permit edits; retained-memory pauses are read-only and reject mutations with 409. SSH, SFTP, file watching, and runtime-only mounts require running compute.

Host-side mounts or a dedicated storage service can provide access while a guest is stopped. Implement authentication, consistency and lifecycle fencing explicitly, especially around retained memory; that is broader work than keeping local disks persistent.

See paused file operations →

Disk ownership already provides full-system persistence

A physical server with local system disks and root access can retain installed tools, configuration and user data. A concrete hardware reference is Hetzner AX42, not a universal fleet design. Redundancy, consistent backups and recovery on replacement hosts require an explicit storage plan.

Source 1 ↗

Powering down is different from releasing reserved hardware

The reference hosting provider continues billing products retained in the account even when powered off. Fleet utilization and scheduler density therefore affect per-agent cost. Owned hardware has a different purchase and operating-cost model; neither implies that idle machine capacity becomes free.

Source 1 ↗

Operate the complete state lifecycle

A fleet must supply the scheduler, sandbox boundary, authenticated file interface, lifecycle coordination and recovery appropriate to its workloads. Host-side tooling can access some guest disks without running the guest, but safe coordination with live or retained process state is an engineering responsibility. A hardware rental quote does not include those runtime services.

Source 1 ↗

Pricing

Compare billing models, then price your workload

CPU allocation, memory, billing minimums, disk, plan fees, and supported workload differ. These meters are not equal-performance machine quotes. Confirm region, resource units, and included allowances before estimating a total.

Self-operated bare-metal sandbox fleet

Pricing reviewed 2026-10-19

Example AX42 physical-server reference
€0.1586/hour; €99 monthly price
Configured displayed reference with one dedicated IPv4; VAT excluded; whole server, not per agent.
Rate source ↗
Example AX42 setup
€49 once off
Separate hardware-provisioning fee, not a universal fleet requirement.
Rate source ↗
Fleet operations and supporting services
Budget separately for the actual design
Retained rented hardware stays billed; include storage/backup, traffic, software and operations.
Rate source ↗

The AX42 reference illustrates one currently priced rented host, not an end-to-end self-hosted sandbox service quote. Location, hardware, setup, IP, redundancy, utilization and operations change the total. Do not derive universal savings or per-agent throughput from this hardware floor.

Check current Self-operated bare-metal sandbox fleet pricing ↗

Sandbox0

$0.015/GiB-hour compute (billed by the second at approximately $0.000004167/GiB-second), plus $0.02/GiB-month of retained incremental RootFS and memory checkpoint data. Shared base images and platform recovery retention are excluded; paused sandboxes accrue storage charges only.

Compute / GiB-hour
$0.015
Retained data / GiB-month
$0.02
Paused compute
$0.00

A 4 GiB agent, 100 running hours

$6.00 compute

4 GiB × 100 hours, billed by the second. Keeping 20 GiB of billable incremental data for a full billing month adds $0.40. CPU is derived from memory by the platform; choose a tier that fits your workload.

Understand storage billing →

Before migrating from Self-operated bare-metal sandbox fleet

  1. Inventory physical capacity, isolation/runtime design, disk placement, backups, recovery, process checkpoints and all operating costs.
  2. Export required installed-system/user state and move compatible Linux agent stages to managed Sandbox0; retain hardware-bound stages separately.
  3. Test isolation requirements, system persistence, online growth, paused files and measured utilization before comparing complete fleet costs.

Common questions

Is Sandbox0 a drop-in Self-operated bare-metal sandbox fleet replacement?

Evaluate API compatibility and product scope. These products serve different core workflows, so a change may require a different architecture.

Does keeping files also keep a running agent?

Filesystem state and volatile process state are separate. Sandbox0 requires explicit compatible memory checkpoints to preserve running processes; its default filesystem resume starts fresh processes.

Sources and review scope

The Boat team's Battleships directory is the discovery source. References below describe the product and plan discussed on this page. Check current availability, regional terms and billing units before migrating. Where the cited material does not establish a capability or price, the comparison identifies that uncertainty.

Battleships research record ↗www.hetzner.com/dedicated-rootserver/ax42/ ↗docs.hetzner.com/general/billing-and-account-management/billing-at-hetzner/billing-system-hetzner/ ↗Sandbox0 RootFS capacity →

Related comparisons

Butter vs Sandbox0 →Cedana vs Sandbox0 →ScitiX Agent Sandbox vs Sandbox0 →Tart Virtualization + Orchard vs Sandbox0 →Agency Tool Company vs Sandbox0 →Agent 37 vs Sandbox0 →