21. Release Notes

21.1. Robin Cloud Native Storage v5.3.14

The Robin Cloud Native Storage (CNS) v5.3.14 release has a new feature, improvements, and a known issue.

Release Date: 18 August 2022

21.1.1. New Features

21.1.1.1. Install Robin CNS on Open-source Kubernetes

Starting from Robin CNS v5.3.14, you can also install Robin CNS on Open-source Kubernetes in the on-premises setup.

21.1.1.2. Install Robin CNS from IBM Catalog

Starting from Robin CNS v5.3.14, you can install Robin CNS from IBM Catalog as a Service on the IBM Managed Kubernetes Service (IKS).

21.1.2. Improvements

21.1.2.1. Option to add Disks as part CNS installation

You can add disks as part of Robin CNS installation on IBM Cloud Kubernetes Service (IKS) and Google Kubernetes Engine (GKE). The following is the option you can use when installing Robin CNS.

[--storage-disks='count:<>,type:<>,size:<>']

21.1.2.2. Install option for DNS Unresolvable Setups

Robin CNS provides the following option for DNS unresolvable setups:

--update-coredns

You can use this option as part of Robin CNS installation for a successful installation.

21.1.3. Known Issues

Reference ID

Description

PP-27689

symptom

In certain scenarios, when you install Robin CNS 5.3.14 from IBM Catalog, the status of the Robin services shows as Active on the IBM Catalog. However, the installation might fail, and you notice the status of the nodes as Notready when you run the robin host list command.

The issue is due to the httpd service not starting. However, this issue does not disturb your IBM IKS services irrespective of the node status.

Apply the following Workaround if you notice this issue.

Workaround

  1. Run the following command to discover the robin-ds pod on which the host is stuck in the NotReady state.

    kubectl get pods -n robinio -l=app=robin -o wide | grep <hostname>

  2. Run the following command to access the robin-ds pod for the host that is in the NotReady state.

    kubectl exec -it <podname> -n robinio -- /bin/bash

  1. Run the following command to check httpd status.

    systemctl status httpd

  1. If the httpd service is down, run the following command to restart.

    systemctl restart httpd

  1. If the httpd service restarts in the previous step, run the following command.

    systemctl restart robin-bootstrap

PP-26360

symptom

Due to the unavailability of requested data on one of the RDVM nodes, a fault event is sent to the master node where stormgr is running. As part of handling the fault, the stormgr might crash if the slice config on RDVM node is at a different version.