Roadmap
What is planned but not yet in Delegaris, in rough order. None of it has a date.
Other places machines live
The resource list, access rules, keep-ons, bookings and approvals do not depend on AWS: each service is a driver behind the same start, stop, reboot and tag actions. Next, each as its own kind of account:
| Platform | How Delegaris would reach it | Stands in for tags |
|---|---|---|
| Microsoft Azure | Subscriptions beside AWS accounts, signed in with a service principal or managed identity. Virtual machines (deallocate and start), SQL databases, App Service plans. | Azure tags |
| VMware vSphere | A vCenter, or a standalone ESXi host, through its REST API. VMware Tools gives a clean guest shutdown. | vSphere tags and categories |
| Proxmox VE | A cluster, through its API with an API token. VMs and containers. | Proxmox tags |
| Microsoft Hyper-V | Hosts or a failover cluster over PowerShell remoting, or System Center VMM where there is one. | The VM's notes field, or a naming convention |
The on-premises three need the container to reach the management network. Keeping a machine on past its schedule works by lifting the schedule tag, so it only means something where a scheduler reads that tag; otherwise Delegaris would need to run the schedules itself, which it does not do today.
Also planned
- Approvals for whole server groups - one request for a group whose members each need approval, instead of refusing the run.
- Kerberos sign-in - no password prompt on domain-joined PCs. Until then, Microsoft Entra ID gives the same experience on hybrid-joined machines.
- Very large estates - fetching each account's inventory on demand and paging the list, for estates beyond a few thousand resources.
- A high-availability build on AWS - the service side is built (see High availability); a tested AWS reference build is not.