0.22.3

Hot Aisle bare metal (experimental)

dstack can now provision Hot Aisle bare metal servers with 8xMI300X in addition to VMs. Bare metal is disabled by default; enable it with bare_metal: true in the backend settings:

projects:
- name: main
  backends:
  - type: hotaisle
    team_handle: hotaisle-team-handle
    creds:
      type: api_key
      api_key: ...
    bare_metal: true

And the bare metal offers become available:

$ dstack offer --backend hotaisle
 #  BACKEND                   RESOURCE TYPE  PRICE
 1  hotaisle (us-michigan-1)  cpu=x86:104 mem=2048GB    bm-mi300x-8    $27.12
                              disk=11
                              gpu=MI300X:192GB:8

Bare metal servers are prepaid for 8 hours. Use a fleet with fixed nodes and delete it when done:

type: fleet
name: hotaisle-bm

nodes: 1
backends: [hotaisle]
instance_types: [bm-mi300x-8]

Within the 8 hours, dstack doesn't force-terminate bare metal servers: the instance stays terminating for about 15 minutes, then dstack stops tracking it and the server must be terminated manually in Hot Aisle. Controlled by DSTACK_FF_HOTAISLE_BARE_METAL_NO_FORCE_RELEASE (enabled by default, 0 disables).

What's Changed

• Add Hot Aisle bare metal support by @peterschmidt85 in #4325
• Correctly support NVIDIA B300 in Vast.ai by @jvstme in #4330
• Update Azure GRID driver to 580.178.04 by @peterschmidt85 in #4332
• [Docs] Update vLLM PD example by @un-def in #4336
• Fix pyright with kubernetes 37 by @r4victor in #4347
• Skip placement groups for AWS Capacity Blocks by @r4victor in #4345
• Warn dstack skill about heredocs split across commands items by @r4victor in #4348
• Bound SSHTunnel control commands with timeouts and BatchMode by @r4victor in #4354
• Use Vultr backend credentials when fetching GPU offers by @peterschmidt85 in #4335
• Reconnect gateway replica SSH tunnels after process exit by @peterschmidt85 in #4355
• Fix gateway recovery after high request load by @peterschmidt85 in #4351

Full Changelog: 0.22.2...0.22.3