@Danp the issue is "resolved" only because we got the host back online so it released the lock.

1. Yes I did try Start on which is method that produced the "The VDI is not available" error
2. no the offline host was not the pool master
3. Yes the VDI's are on shared iSCSI storage (truenas)

And I want to note that OTHER VM's that were not on that failed host were able to restart and on any other node without an issue.

I was even able to delete old snapshots without an issue so clearly the storage was online and available.

It is deeply concerning that we were unable to get the VMs running again and only coincidental that were were able to re-seat the ram module and get it back online -- once it rejoined the pool the locked VM started without any issues on another host, so something was clearly locking it down even though it was fully stopped and visible.

FYI I was unable to disconnect the SR for the offline host as well:

pbd.disconnect
{
  "id": "9355e1d9-164c-76a8-6ab7-aba7a9793b31"
}
{
  "code": "HOST_OFFLINE",
  "params": [
    "OpaqueRef:769bbac2-3fc3-4761-91d1-51fa8468fc72"
  ],
  "task": {
    "uuid": "3cb1c5e9-e7cb-83b5-e448-ccddc3c4fddf",
    "name_label": "Async.PBD.unplug",
    "name_description": "",
    "allowed_operations": [],
    "current_operations": {},
    "created": "20250919T14:58:59Z",
    "finished": "20250919T14:58:59Z",
    "status": "failure",
    "resident_on": "OpaqueRef:32345a51-c14d-4b3c-ae6e-e411c3c8808c",
    "progress": 1,
    "type": "<none/>",
    "result": "",
    "error_info": [
      "HOST_OFFLINE",
      "OpaqueRef:769bbac2-3fc3-4761-91d1-51fa8468fc72"
    ],
    "other_config": {},
    "subtask_of": "OpaqueRef:NULL",
    "subtasks": [],
    "backtrace": "(((process xapi)(filename ocaml/xapi/message_forwarding.ml)(line 124))((process xapi)(filename ocaml/xapi/message_forwarding.ml)(line 160))((process xapi)(filename lib/xapi-stdext-pervasives/pervasiveext.ml)(line 24))((process xapi)(filename ocaml/xapi/rbac.ml)(line 205))((process xapi)(filename ocaml/xapi/server_helpers.ml)(line 95)))"
  },
  "message": "HOST_OFFLINE(OpaqueRef:769bbac2-3fc3-4761-91d1-51fa8468fc72)",
  "name": "XapiError",
  "stack": "XapiError: HOST_OFFLINE(OpaqueRef:769bbac2-3fc3-4761-91d1-51fa8468fc72)
    at Function.wrap (file:///usr/local/lib/node_modules/xo-server/node_modules/xen-api/_XapiError.mjs:16:12)
    at default (file:///usr/local/lib/node_modules/xo-server/node_modules/xen-api/_getTaskResult.mjs:13:29)
    at Xapi._addRecordToCache (file:///usr/local/lib/node_modules/xo-server/node_modules/xen-api/index.mjs:1073:24)
    at file:///usr/local/lib/node_modules/xo-server/node_modules/xen-api/index.mjs:1107:14
    at Array.forEach (<anonymous>)
    at Xapi._processEvents (file:///usr/local/lib/node_modules/xo-server/node_modules/xen-api/index.mjs:1097:12)
    at Xapi._watchEvents (file:///usr/local/lib/node_modules/xo-server/node_modules/xen-api/index.mjs:1270:14)"
}

Here you go:

vm.start { "id": "838dfee6-4b1c-0d12-7291-c128e663cb62", "bypassMacAddressesCheck": false, "force": false } { "code": "SR_BACKEND_FAILURE_46", "params": [ "", "The VDI is not available [opterr=['HOST_OFFLINE', 'OpaqueRef:769bbac2-3fc3-4761-91d1-51fa8468fc72']]", "" ], "call": { "duration": 5670, "method": "VM.start", "params": [ "* session id *", "OpaqueRef:2a8b259d-5a3b-44cc-9e09-c938b1fb4b33", false, false ] }, "message": "SR_BACKEND_FAILURE_46(, The VDI is not available [opterr=['HOST_OFFLINE', 'OpaqueRef:769bbac2-3fc3-4761-91d1-51fa8468fc72']], )", "name": "XapiError", "stack": "XapiError: SR_BACKEND_FAILURE_46(, The VDI is not available [opterr=['HOST_OFFLINE', 'OpaqueRef:769bbac2-3fc3-4761-91d1-51fa8468fc72']], ) at Function.wrap (file:///usr/local/lib/node_modules/xo-server/node_modules/xen-api/_XapiError.mjs:16:12) at file:///usr/local/lib/node_modules/xo-server/node_modules/xen-api/transports/json-rpc.mjs:38:21 at runNextTicks (node:internal/process/task_queues:60:5) at processImmediate (node:internal/timers:454:9) at process.callbackTrampoline (node:internal/async_hooks:130:17)" }

JUST ran into this today -- was able to FINALLY get the dead node back online (RAM module had gone off and re-seated it) -- but OMG this is terrible situation -- we were able to get vms to an "offline" status but were wholly unable to get them to start on any other node!

Surely there must be some way from the command line to unstick this?