hosted.ai v2.4.1 - scalable neocloud storage, enhanced GPUaaS workflows + auditing

Version 2.4.1 of the hosted.ai neocloud platform is available now. This release adds more storage scalability for GPU cloud operators, and a number of UI improvements to streamline workflows, enhance usability, and improve visibility and auditing.

‍

What's new in hosted·ai v2.4.1:

‍

‍1. Improved Ansible playbook visibility in GPUaaS pods

The hosted·ai platform has a flexible recipe system. It allows GPUaaS pod instances to be customized with Ansible playbooks for automated installation of AI models, frameworks and applications from the hosted·ai library.

In hosted·ai v2.4.1, we’ve enhanced visibility of recipe execution by introducing role-based execution logs to the User Panel.

• Standard users (e.g. customers) can view high-level service configuration events - recipe started, completed, or failed - and can automatically report failures to system administrators with a single click
  ‍
• Administrators have access to detailed Ansible execution logs for troubleshooting, including instance GPU, CPU, RAM and resource pool details
  ‍
• Execution logs are retained before Kubernetes TTL clean-up, or instance decommissioning, to ensure a permanent audit trail

‍

What this means:
There is now a permanent audit trail of recipe activity across the platform, visible at the correct depth by user role.

‍

‍2. Multiple storage volume groups per region, automatic allocation

Hosted·ai v2.4.1 now enables administrators to register multiple storage nodes per region. Previously only a single volume group per region was allowed. Existing single volume group regions continue to operate without any migration or reconfiguration required.

• The platform automatically selects the optimal volume group for each storage request based on absolute free space - no user or admin input is needed
  ‍
• It uses best-effort instance-storage colocation to minimize latency
  ‍
• Admins now have region- and node-level storage utilization dashboards, with team/volume drill-down, and a standardized shared volume creation flow that is consistent with other platform creation workflows
  ‍

Why this matters:
‍This improves regional storage scalability, allowing GPU cloud administrators to expand storage online without disrupting existing infrastructure or replacing resources. It reduces single points of storage failure while maintaining full backwards compatibility with existing storage configurations.

‍

3. Floating IP management for GPUaaS nodes from the UI

In hosted·ai 2.4.1, administrators can view, create, edit and delete floating IP mappings per region from the UI, instead of the previous SSH + CLI configuration process.

• There is a new Floating IPs sub-tab in the Admin Panel, under Infrastructure > Nodes > Accelerator
  ‍
• IP addresses are auto-discovered from per-node re-scan results, or can be entered manually with input validation (well-formed IPv4, 1:1 mapping enforcement)
  ‍
• The hosted·ai cluster orchestrator remains the authoritative source: the Admin Panel reconciles its local state on each sync

‍

Why this is useful:
‍It reduces the time needed to assign a floating IP from minutes to seconds, eliminates dependency on manual CLI operations for routine IP management, and minimizes the risk of configuration errors caused by unvalidated command-line entries.

‍

4. UI improvements to navigation and GPUaaS metrics

This version also adds a couple of quality-of-life UI improvements to navigation and GPUaaS pool metrics:

‍

• The User Panel has improved "Go back" behaviour in the instance detail page, with the breadcrumb and a new “back” button adapting to the user's entry point
  ‍
• GPUaaS pool metrics are now correctly scoped per instance creation timestamp: now re-subscribing to a pool starts from a clean baseline, while historical instance usage can also be displayed in the UI, labelled with the GPUtype/pool name that was in effect when the instance was previously active, for an easier-to-understand audit trail
  ‍

The benefits:
‍These improvements reduce potential friction and confusion for customers, and reduce support requests for GPU cloud operators, by providing a more intuitive and consistent user experience.

‍

‍Next steps

This release also includes many smaller features and fixes based on customer feedback.

• For the full list of changes, see the v2.4.1 release notes
  ‍
• To upgrade from previous versions to hosted·ai v2.4.1, please contact your account manager or our customer success team.
  ‍
• If you're new to hosted·ai, get in touch for a demo and we'll walk/talk you through the platform. Thanks!