File tree Expand file tree Collapse file tree 2 files changed +29
-3
lines changed Expand file tree Collapse file tree 2 files changed +29
-3
lines changed Original file line number Diff line number Diff line change @@ -243,10 +243,34 @@ the host.
243
243
244
244
The cloudstack agent uses the ``gpudiscovery.sh `` script to discover the GPU
245
245
devices on a KVM host. The script is located in the
246
- ``/usr/share/cloudstack-common/scripts/vm/ `` directory on the host.
246
+ ``/usr/share/cloudstack-common/scripts/vm/ `` directory on the host. The script
247
+ relies on the ``lspci `` & ``xmlstarlet `` command to discover the GPU devices
248
+ and their status on the host. So, for the discovery to be successful, the
249
+ ``lspci `` & ``xmlstarlet `` commands needs to be installed on the host.
250
+
251
+ .. parsed-literal ::
252
+
253
+ dnf install pciutils xmlstarlet
254
+
255
+ .. parsed-literal ::
256
+
257
+ sudo apt install pciutils xmlstarlet
247
258
248
259
.. note ::
249
- The script can be run manually to debug the discovery of the GPU devices on a host.
260
+ The following table shows the compatibility matrix for NVIDIA vGPU types with CloudStack:
261
+
262
+ .. cssclass :: table-striped table-bordered table-hover
263
+
264
+ =============================== ================== =======================
265
+ NVIDIA vGPU Type VFIO Framework Supported in CloudStack
266
+ =============================== ================== =======================
267
+ Legacy: SR-IOV not supported `mdev ` Yes
268
+ SR-IOV supported `mdev ` Yes
269
+ SR-IOV supported `Vendor specific ` No
270
+ Multi Instance GPU No
271
+ =============================== ================== =======================
272
+
273
+ The script can also be run manually to debug the discovery of the GPU devices on a host.
250
274
251
275
.. parsed-literal ::
252
276
Original file line number Diff line number Diff line change @@ -304,7 +304,9 @@ To create a new compute offering:
304
304
This is applicable only for KVM hypervisor.
305
305
306
306
- **GPU Display **: Whether to use the GPU device attached to the guest instance for display.
307
- This is applicable only for KVM hypervisor.
307
+ This is applicable only for KVM hypervisor. Depending on the OS and display configuration,
308
+ the user might need to set ``video.hardware `` to ``none `` in the instance's settings to
309
+ use CPVM for display.
308
310
309
311
- **Public **: Indicate whether the compute offering should be
310
312
available to all domains or only some domains. Choose Yes to make it
You can’t perform that action at this time.
0 commit comments