-
-
When starting a glusterfs pod from the glusterfs DaemonSet (or when restarting those pods), net.bridge.bridge-nf-call-iptables is set to 0 on the host machine, which wreaks havoc on that Kubernetes no…
-
These are just some comments on the `images:` list in the configuration template
https://github.com/Azure/az-hop/blob/3b943a965e72d8757d1b1ec36bdc1e15e4074ebb/config.tpl.yml#L258-L339
1. It wo…
-
**We tested GDR in baremetal with nccl successfully as belows (pcie acs has been prohibited):**
```
instance-ubm6ko9y:163876:163916 [4] NCCL INFO Channel 02/0 : 13[b2000] -> 4[b1000] [receive] via…
-
After trying ./configure and `bazel build -c opt //tensorflow_networking/mpi:all` I get the error
```
ERROR: Skipping '//tensorflow_networking/mpi:all': while parsing '//tensorflow_networking/mpi:…
-
Hello! Did BytePS implement multiple NICs internally?
-
Hi, All:
when i Run nvdla_runtime --loadable kmd/PDP/PDP_L0_0_small_fbuf in nv_small fpga , i got error like this Invalid dst_data.mem_type: 2048.
= Run PDP/PDP_L0_0_small_fbuf
creating new…
-
Hi,
I am trying to setup ntrdma(https://github.com/ntrdma/ntrdma-ext) with Intel Skylake b2b configuration on Linux 4.14.178.
Following is how my memory windows are setup:
1) MW[0] (BAR23): dma_a…
-
If an argument for an action is bitwise copy able (like `std::vector` for `T` being bitwise copy able) it should be possible to avoid to copy it into (and out of) the serialization buffer. Investigate…
-
### Describe the bug
ROCm related unit test failed
see [rocm.log](https://github.com/openucx/ucx/files/9076538/rocm.log)
### Setup and versions
- CentOs stream 8
- rdma-core-55mlnx-37-1.5510…