20. Release Notes

20.1. Robin Cloud Native Storage v5.4.18-281

The Robin CNS v5.4.18-281 Release Notes document provides information about upgrade paths, improvements, fixed issues, and one known issue.

Release Date: January 12, 2026

20.1.1. Upgrade Paths

The following are the supported upgrade paths for Robin CNS v5.4.18-281:

  • Robin CNS v5.4.16-105 to Robin CNS v5.4.18-281

  • Robin CNS v5.4.16-167 to Robin CNS v5.4.18-281

Note

  • After upgrading to Robin CNS v5.4.18-281, if you are using the Robin Client outside the robincli Pod, you must upgrade to the latest version of the Robin Client.

  • If you have installed Robin CNS with the skip_postgres_operator parameter to use the Zalando PostgreSQL operator, then you must first upgrade the Zalando PostgreSQL operator to v1.11.0 or later before upgrading to Robin CNS v5.4.18-281.

  • After upgrading from any supported Robin CNS version to Robin CNS v5.4.18, certificates will be automatically renewed.

20.1.2. Improvement

20.1.2.1. Readiness probe support for IOMGR service

Starting with Robin CNS v5.4.18, Robin CNS introduces Readiness probe support for the IOMGR service. These probes verify that both RIO and RDVM services are operational and ready to serve I/O operations. When both services are ready, then only Robin CNS marks the IOMGR service as ready. For more information about Readiness probe, see Readiness probe.

20.1.2.2. Improved Stability and Performance for Windows VMs

In Robin CNS, I/O operations can experience some latencies during storage device initialization (e.g., after a node restart or during recovery). This impacts application response and causes Windows VMs to freeze.

The following improvements are made in the Robin CNS to solve this issue:

  • Ensure that the volume mount and volume slice leader are on the same node for optimal I/O performance.

  • Fixed the Block map cache assertion to prevent IOMGR restarts.

  • Improved the Block map load time.

  • Optimized Garbage Collection (GC) for volumes with 512 block size.

  • Avoid IOMGR restarts during Robin master Pod failovers.

20.1.3. Fixed Issues

Reference ID

Description

RSD-9809, RSD-10087

The out-of-sync issue with Patroni Pods that led to a Robin Service Outage is fixed.

RSD-9861, RSD-9886, RSD-9944

The Remote Procedure Call (RPC) client is dropping pending I/O requests without processing the received response. Because of these pending I/O requests, the virtual machine instance (VMI) Pod cannot reconcile the state and shows the following error:

unknown error encountered sending command SyncVMI: rpc error: code = DeadlineExceeded desc = context deadline exceeded

The issue of the VM instance hitting the SyncVMI reconciliation error is fixed.

RSD-9882

The issue of IOMgr failing to restart the volume remount operation due to flaky Kubernetes API service after the StorMgr recovers from network partition is fixed.

RSD-9867

The issue of creating a thick clone volume from its snapshot getting stuck in the hydrating state is fixed.

RSD-9775

The Robin CNS v5.4.16 installation failed because the robin-patroni-pre-install-hook job did not execute successfully, leaving Patroni stale data. This is caused by one of its Pods stuck in the CreateContainerError state during the installation. This issue is fixed.

RSD-9603

When a node with an active master Pod is rebooted, in the kubectl describe robinclusters command output, the Phase field shows the correct status of the Robin CNS cluster as NotReady, but the phase_reason field is empty. This issue is fixed.

RSD-9518

The Sherlock tool is incorrectly reporting a node as down. This issue is fixed by removing the SSH service from its diagnostic checks.

RSD-9422

In a rare scenario, when a thick clone volume’s hydration is in progress, if that clone volume’s PVC is deleted, the stormgr-server goes down, and as a result another thick clone volume’s hydration gets stuck. This issue is fixed.

RSD-7375

A volume clone creation with a larger size than the source volume is not working. This issue is fixed.

RSD-9165

When a node is removed and added back to the cluster, Robin Patroni Pods are stuck in the Pending state on that node, as required labels were not added to the replaced node. This issue is fixed.

RSD-9654

In a rare scenario, IO operations hang up on the Robin volume slice due to a network blip. This issue is fixed.

RSD-8083

To ensure quicker failover, tasks related to device slice leader change have been optimized. This is especially beneficial during node reboots in environments with nodes containing many large devices, as following slice operations can now finish faster.

RSD-10021

The Patroni cluster might go down if some of the nodes are cordoned in a rare scenario. This issue is fixed.

RSD-10248

When a node abruptly powers off, Pods that use persistent volumes are rescheduled to other nodes. However, in some cases, some of these Pods fail to start on the new nodes with a Multi-Attach error because the volume remains exclusively attached to the powered-off node. When you run kubectl get events, you might see the following error:

Multi-Attach error for volume … Volume is already exclusively attached to one node and can’t be attached to another.

This issue is fixed.

RSD-10342

The issue of noisy logs generated for the bmap_update_entry function in the IOMGR logs is fixed.

RSD-10348

The issue of block checksum mismatch occurring after upgrading to Robin CNS v5.4.18 is fixed.

RSD-10357

When creating a cloned volume on a device with a write unit different than the device where the parent volume is provisioned, the IOMGR service is restarting due to an assert. This issue is fixed.

RSD-10530

The issue of PersistentVolumeClaim (PVC) creation failing when provisioning a Persistent Volume (PV) due to the unavailability of the Kubernetes API server is fixed.

RSD-10565

The malformed JSON as part of one of the metrics payload is now fixed.

PP-38537

After deleting a backup, unregistering a storage repo fails with the following error message:

Storage repo is associated with volume group

This issue is fixed.

PP-39836

The issue of I/O operations not failing during device initialization or redo replay operations, causing unnecessary I/O latency for volumes, is fixed.

RSD-10242

The Robin worker Pod is crashing due to the empty /etc/robin/rcm/config.ini file. This issue is fixed.

20.1.4. Known Issue

Reference ID

Description

PP-40480

Symptom

In rare scenarios, you might observe that one of the Pods is stuck in the ContainerCreating state, and the kubectl describe pod command shows the following volume mount error:

Failed to mount volume pvc-d16fa6b1-5bcb-4c69-805d-ab4df9018cee: Node <default:vnode-87-237> has mount_blocked STORMGR_NODE_BLOCK_MOUNT. No new mounts are allowed.

Workaround

Bounce the worker Pod running on the affected node.

20.1.5. Technical Support

Contact Technical support for any assistance.

20.2. Robin Cloud Native Storage v5.4.18-301

The Robin CNS v5.4.18-301 Release Notes document provides information about upgrade paths, fixed issues, and known issues.

Release Date: July 27, 2026

20.2.1. Infrastructure Versions

The following software applications are included in this CNS release:

Software Application

Version

Patroni

4.0.4

PostgreSQL

15.10

Zalando PostgreSQL Operator

1.14.0

Spilo

17:4.0-p2

20.2.2. Upgrade Paths

The following are the supported upgrade paths for Robin CNS v5.4.18-301:

  • Robin CNS v5.4.16-166 to Robin CNS v5.4.18-301

  • Robin CNS v5.4.18-278 to Robin CNS v5.4.18-301

Note

  • After upgrading to Robin CNS v5.4.18-301, if you are using the Robin Client outside the robincli Pod, you must upgrade to the latest version of the Robin Client.

  • If you have installed Robin CNS with the skip_postgres_operator parameter to use the Zalando PostgreSQL operator, then you must first upgrade the Zalando PostgreSQL operator to v1.11.0 or later before upgrading to Robin CNS v5.4.18-301.

  • After upgrading from any supported Robin CNS version to Robin CNS v5.4.18, certificates will be automatically renewed.

20.2.3. Fixed Issues

Reference ID

Description

RSD-11147

When the robin-patroni-pre-install-hook job fails on a node during the Robin CNS installation, the has-run-patroni-pre-install-job=true label is incorrectly applied. This prevents the pre-install-hook Pod from being scheduled on that node during subsequent attempts, causing the installation to fail. This issue is fixed.

RSD-11042

The issue of the standby robin_master Pod failing its readiness probe and becoming unhealthy due to a mismatch between the in-memory phase state and the value stored in the /etc/robin/robin_master/phase directory is fixed.

RSD-11474, RSD-11238

For a multi-attach volume, when one of the nodes became faulty, apart from the faulty node, the volume was incorrectly unmounted from other healthy nodes. This issue is fixed.

RSD-1016

The issue where VolumeAttachment jobs could get stuck due to a race condition and lock contention within the job manager and prevent volume mount and unmount operations from progressing is fixed.

RSD-10603

During the Robin CNS upgrade, the IOMgr server failed to start because of the empty /etc/consul/client/checks.json file on the node. This issue is fixed.

RSD-11233

The issue of the Robin cluster remaining in the NotReady state when a host gets stuck in the INIT_FAILED state due to the Kubernetes API server flakiness is fixed. Now, the Robin CNS operator will retry in case of failure.

RSD-12074

The issue where IOMgr server exhausted all retries and entered an infinite restart loop when consul was unavailable is fixed. This occurred because the IOMgr server attempted to register before consul was ready to handle registration requests. The IOMgr server now remains active and waits for consul to become available, ensuring the cluster recovers automatically once consul is restored.

RSD-11044

The priorityClass for the robin-tcmuproxy DaemonSet was not set to critical. This issue is fixed. Now, the priorityClass is set to critical.

RSD-12098

The issue where the robin-master Pod failed to recover automatically after a network partition is fixed. This failure occurred because the robin-server failed to start due to restarts of all kube-apiserver instances, which caused the REST interface to remain unavailable.

RSD-11041

The issue of Robin cluster resources reporting the NotReady status for an extended period of time is fixed. The backoff time is adjusted to display accurate cluster status.

20.2.4. Known Issue

Reference ID

Description

PP-43541

Symptom

In a rare scenario, a robin-worker Pod might be stuck in a CrashLoopBackOff state, and Pod logs show the error related to configparser.

configparser.MissingSectionHeaderError: File contains no section headers. Setting up cluster config failed.

This occurs because a partial and incomplete config.ini file is generated in the /home/robinds/etc/robin/rcm/ directory, which causes the Robin agent bootstrap process to fail.

Workaround

To solve this issue, perform the following steps:

  1. Truncate the partially written config.ini file on the affected robin-worker Pod:

    # truncate --size 0 /home/robinds/etc/robin/rcm/config.ini

  2. Bounce the affected robin-worker Pod:

    # kubectl delete pod <robin_worker> -n robinio

PP-40480

Symptom

In a rare scenario, you might observe that one of the Pods is stuck in the ContainerCreating state, and the kubectl describe pod command shows the following volume mount error:

Failed to mount volume pvc-d16fa6b1-5bcb-4c69-805d-ab4df9018cee: Node <default:vnode-87-237> has mount_blocked STORMGR_NODE_BLOCK_MOUNT. No new mounts are allowed.

Workaround

Bounce the robin-worker Pod running on the affected node:

# kubectl delete pod <robin_worker> -n robinio

PP-43857

Symptom

When upgrading to Robin CNS v5.4.18-301, the robin-worker Pod might get stuck in the Terminating state due to pending I/O operations on the node. This occurs because the IOMgr Pod fails to start as it is waiting for a healthy consul-client , while the consul-client cannot start until the robin-worker Pod is stuck in the Terminating state.

Workaround

Delete the affected robin-worker Pod with --force option:

# kubectl delete pod <robin_worker> --force -n robinio

20.2.5. Technical Support

Contact Technical support for any assistance.