Doron KG
Doron KG
Updating status, we've upgraded the NVIDIA GPU Operator to v22.9.2, while upgrading the NVIDIA GPU Driver to v525.85.12. The v22.9.2 installs Driver v525.60.13 by default, in order to install v525.85.12...
**UPDATE:** We've upgraded to NVIDIA GPU Operator v23.3.2 with GPU Driver v535.104.12 (recommended, not default). The benchmark resulted in `train.total_ips` average of ~2600 ips in each iteration.
I suggest renaming `--print-reason` flag to `--show-reason` in similarity to kubectl flags.
Hi @isindir, In order for the new feature to be completed, several additional files should be modified accordingly. Take a look at https://github.com/yonahd/kor/blob/main/CONTRIBUTING.md#repository-structure.
Hi @matazr, although the PR was merged, there're several disparities with README. The `AppVersion` still points to v0.3.4, and many keys refer to `prometheusExporter.service` rather than `prometheusExporter.serviceMonitor`.
> Helm chart I think needs extra changes: > > * at the moment it is not possible to configure `kor` in a deployment for example to automatically remove any...
The issue results from the `retrieveUsedClusterRoles()` function that is called in `pkg/kor/clusrterroles.go`. ``` func retrieveUsedClusterRoles(clientset kubernetes.Interface, filterOpts *filters.Options) ([]string, error) { //Get a list of all namespaces namespaceList, err :=...
> > @hrytsai this worked fine in 0.3.7? > > Nope, the same problem Let's split it into two cases: **1)** The `--resources` flag was only introduced in v0.3.8, so...
> Hi > > Thanks for explanation, we have case 2. > > But problem is that we cannot provide SA\User for get, list, watch namespaces permission in multi tenancy...