-
# Required information
* Distribution: Ubuntu Asahi
* Distribution version: 24.04
* The output of "incus info":
incus info
config: {}
api_extensions:
- storage_zfs_remove_snapshots…
-
I noticed a performance regression in OSU benchmark (OpenMPI with UCX and HCOLL) when using HPCX 2.17.1 compared to 2.14. It is due to the fact that now the `UCX_PROTO_ENABLE=y` by default. Setting it…
-
## CVE-2019-17075 - High Severity Vulnerability
Vulnerable Library - linuxlinux-4.19.236
The Linux Kernel
Library home page: https://mirrors.edge.kernel.org/pub/linux/kernel/v4.x/?wsslib=linux
Foun…
-
Dears,
First I would like to thank you for the great work you have done to help using Dask+GPU.
The LocalCUDACluster is really great and makes things much more simple.
Actually I have been able…
-
I try to use your example and use Service to proxy the glusterfs pod, so that we does not need to use host network and care about where the Pod is running on. The yaml is like:
```
---
kind: Service…
-
### Describe the bug
Our application is giving a segfault when setting `UCX_TLS=self,sm,ud_mlx5`. Everything is working fine when replacing `ud_mlx5` with `ud_verbs` or `ud`. Interestingly, setting `…
-
I compiled 1.6.1 with GCC/7.3.0-2.30 and CUDA/9.2.88. When I try to run the `ucx_perftest` on our InfiniBand cluster with `-m cuda` or `-m cuda-managed`, I see the following segfault:
```
./ucx_pe…
-
Hi Ravindra,
1. Lovely documentation. I showed it to class as an example of the standard I want to see in general.
2. No automated tests. You've made a start, but a feeble one. I expect better!
3. No…
-
I came across the PR for Open MPI that added `shmemx_malloc_with_hint`, implemented in terms of `uct_ib_md_alloc_device_mem` (https://github.com/open-mpi/ompi/pull/6641). I would like to give this shm…
-
### Describe the bug
We have a couple of ARM (Thunder) nodes to play with before we get our A64fx cluster. The nodes have OFED5 on them and mlx5 devices. I get free passive progress from OFED4 + …