Cisco Nexus Dashboard Release Notes, Release 2.1.2

March 13, 2023

Updated the recommended CIMC version to 4.2(3b).

November 23, 2022

Release 2.1.2f became available.

September 15, 2022

Updated “Sites per cluster” scale in the “Verified Scalability Limits” section.

August 17, 2022

Removed CSCvw52468 and CSCvy33462 from the “Known Issues” section, because they are not applicable for this release.

August 1, 2022

Updated the recommended CIMC version to 4.2(2a).

April 28, 2021

Issue CSCwa45116 was incorrectly listed as Open instead of Resolved. Moved the information into the “Resolved Issues” section.

March 15, 2021

Additional open issue CSCwb18594.

December 19, 2021

Release 2.1.2d became available.

Ability to Onboard Cloud APIC Sites from Google Cloud Platform

You can now add Cisco Cloud APIC sites that are managing Google Cloud Platform to your Nexus Dashboard cluster.

For more information, see Cisco Nexus Dashboard User Guide.

Multi-Factor Authentication (MFA) Using Cisco Duo

This release adds support for Multi-Factor Authentication (MFA) for your Nexus Dashboard using Okta and Cisco Duo.

For more information about cluster sizing, see Nexus Dashboard Cluster Sizing tool.

For more information about deploying a cluster, see Cisco Nexus Dashboard Deployment Guide.

For more information about deploying services, see the service-specific documentation.

Pre-installation of Nexus Dashboard Insights and Orchestrator Services When Ordering Physical Nodes

If you are ordering physical Nexus Dashboard servers, you now have the option to choose Nexus Dashboard Insights and Nexus Dashboard Orchestrator services to be pre-installed on the hardware before it is shipped to you. Note that if you are deploying the virtual or cloud form factors of the Nexus Dashboard, there are no changes to service installation and you will need to deploy the services separately after the cluster is ready.

For more information, see the Nexus Dashboard Ordering Guide.

CSCvy82547

To do upgrade ND, you need to:

1. Disable the installed apps, and

2. Validate health of the  cluster before proceeding with upgrade of the a node (using acs health)

Otherwise, you can end up in situations where some kafka topics will have no leader.

2.1.2d and later

CSCvy97888

Unable to delete a security domain from site edit page.

2.1.2d and later

CSCvz42210

After completing 2-node RMA procedure, new nodes' serial numbers are overwritten with old nodes' serial numbers, causing cluster to become unhealthy.

2.1.2d and later

CSCvz50040

"kubectl get pods -A" reports a pod in "ContainerCreating" state.

-------------------------------------------------------------------------------------

nodemgr                    pod/nodeagent-tbh29                                 0/1     ContainerCreating   0          84m     <none>           ute11-nd3   <none>           <none>

"kubectl describe pod/nodeagent-tbh29 -n nodemgr", will have Events that look like:

---------------------------------------------------------------------------------------------------------------

Events:

  Type     Reason       Age                   From                Message

  ----     ------       ----                  ----                -------

  Warning  FailedMount  52m                   kubelet, ute11-nd3  Unable to attach or mount volumes: unmounted volumes=[logs], unattached volumes=[network-config localdb system-version default-token-h5npm config kms logs cloud-config]:

  Warning  FailedMount  27m (x2 over 72m)     kubelet, ute11-nd3  Unable to attach or mount volumes: unmounted volumes=[logs], unattached volumes=[kms logs cloud-config network-config localdb system-version default-token-h5npm config]:

  Warning  FailedMount  22m (x4 over 81m)     kubelet, ute11-nd3  Unable to attach or mount volumes: unmounted volumes=[logs], unattached volumes=[default-token-h5npm config kms logs cloud-config network-config localdb system-version]:

  Warning  FailedMount  8m13s (x44 over 83m)  kubelet, ute11-nd3  MountVolume.SetUp failed for volume "nodemgr-log-nodemgr" : mount command failed, status: Failure, reason: exit status 32

Keywords to look for here are:

 - unmounted volumes=[logs]

 - MountVolume.SetUp failed for volume "nodemgr-log-nodemgr"

2.1.2d and later

CSCvx93124

You see a message like:

  [2021-04-13 13:48:20,170] ERROR Error while appending records to stats-6 in dir /data/services/kafka/data/0 (kafka.server.LogDirFailureChannel)

  java.io.IOException: No space left on device

2.1.2d and later

CSCvz64288

Upgrade from 2.0.2h to 2.1 fails with the following message:

install/1-atomix-install

atomix extract failed

This indicates that the 2.1 ISO was not copied completely across all nodes, and there was an IO error while trying to copy files from that ISO.

2.1.2d and later

CSCvz57307

Upload of firmware image to prepare for upgrade is extremely slow.

2.1.2d and later

CSCwb18594

When trying to add a site into Nexus Dashboard, if the password has an '&' the addition of the site fails and stays in an uknown state. With the following error message:

"Site not available, Verify input:Response error:401 Unauthorized {\"totalCount\":\"1\",\"imdata\":[{\"error\":{\"attributes\":{\"code\":\"401\",\"text\":\"User credential is incorrect - FAILED local authentication\"}}}]}"

2.1.2d and later

CSCvz46957

When worker nodes are added, some workers may get stuck in 'Discovering' state.

2.1.2d

CSCvz87060

Pods are stuck in "ContainerCreating" state and kubectl description of the pods will show the following errors:

Warning  FailedCreatePodSandBox  4m20s (x268 over 91m)  kubelet, nd-node3  (combined from similar events): Failed create pod sandbox: rpc error: code = Unknown desc = failed to create pod network sandbox k8s_installer-659f7846db-sc8np_installer_a4b8b47a-affd-4a87-9372-3865e82e8160_0(e3b0897ce954bb9df4a29d972d44c544ff10636f4c91a60ec55c577a5534d9e8): Multus: [installer/installer-659f7846db-sc8np]: error adding container to network "oob": delegateAdd: error invoking conflistAdd - "oob": conflistAdd: error in getting result from AddNetworkList: failed to get ippool, error: failed to get ippool, error: no such object &{10.1.1.105 {169.254.1.128 ffffff80}}

NOTE: If you do not see the "failed to get ippool" error, this is not the same bug, and DO NOT APPLY WORKAROUND specified here.

2.1.2d

CSCwa26117

Although ND has two PSUs, the number of PSU is shown as one on "System resources -> Nodes -> Click on a node row -> Expand the node view -> Overview and Hardware Resources".

2.1.2d

CSCwa47299

This bug has been filed to evaluate the product against the following vulnerability in the Apache Log4j Java library disclosed on December 9, 2021

CVE-2021-44228: Apache Log4j2 JNDI features do not protect against attacker controlled LDAP and other JNDI related endpoints. Cisco is currently investigating impact.

For more information, see Vulnerability in Apache Log4j Library Affecting Cisco Products: December 2021.

2.1.2d

CSCwa45116

A user may not be able to log in if they have 'approver' or 'deployer' role

2.1.2d

CSCvy85865

One a large cluster the consolidated tech support file can become too large to fit into the file system. In this case you will notice that tech support collection failed. On this failure, you may notice that the tsctrl pod has restarted which can be observed using: "kubectl get pods -n ts"

2.1.2d

CSCvy62110

For Nexus Dashboard nodes connected to Catalyst switches packets are tagged with vlan0 even though no VLAN is specified. This causes no reachability over the data network. In this case, 'switchport voice vlan dot1p' command must be added to the switch interfaces where the nodes are connected.

CSCvw39822

On power cycle system lvm initialization may fail on due to a slowness in the disks.

CSCvw48448

Upgrade fails and cluster is in diverged state with one or more nodes on the target version.

CSCvw57953

When the system is being recovered with a clean reboot of all nodes, the admin login password will be reset to the day0 password that is entered during the bootstrap of the cluster.

CSCvw70476

When bringing up ND cluster first time, all three master nodes need to join Kafka cluster before any master node can be rebooted. Failing to do so, 2 node cluster doesn't become healthy as Kafka cluster requires 3 nodes to be in Kafka cluster first time.

CSCvx89368

After ND upgrade, there will be still pods belonging to the older version running on the cluster.

For example, in this case upgrade was from 2.0.1.27 to 2.0.1.36.

After the upgrade, running following command gives:

node1# kubectl get pods -n kube-system -o yaml | grep image: | grep 2.0.1.27

      image: infra/ui:nd-2.0.1.27-e881b96b5

      image: infra/ui:nd-2.0.1.27-e881b96b5

      image: infra/ui:nd-2.0.1.27-e881b96b5

      image: infra/ui:nd-2.0.1.27-e881b96b5

      image: infra/ui:nd-2.0.1.27-e881b96b5

      image: infra/ui:nd-2.0.1.27-e881b96b5

node1# acs version

Nexus Dashboard 2.0.1.36

Clearly the ND nodes have completed upgrade, but some services are showing older version.

CSCvx98282

Pods in pending state for a long period upon restart. These pods are usually stateful sets that require specific node placement and capacity must be available on the specific node they are first scheduled. This happens when multiple applications are installed on the same ND cluster and the ND capacity overloaded.

CSCvu21304

Intersight device connector connects to the Intersight over the Cisco Application Services Engine Out-Of-Band Management.

Nodes in a physical cluster

3 master nodes
4 worker nodes
2 standby nodes

Nodes in a virtual cluster (ESX)

3 master nodes
3 worker nodes
2 standby nodes

Nodes in a virtual cluster (KVM)

3 master nodes

Nodes in a cloud cluster (AWS or Azure)

3 master nodes

Sites per cluster

100 for Nexus Dashboard and Nexus Dashboard Orchestrator, see Nexus Dashboard Orchestrator Verified Scalability Guide for details and limitations.

4 for Nexus Dashboard Insights

Admin users

50

Operator users

1000

Service instances

4

API sessions

2000 for Nexus Dashboard and Nexus Dashboard Orchestrator

100 for Nexus Dashboard Insights

Login domains

8

Clusters connected via multi-cluster connectivity for single pane of glass experience

4

Sites across all clusters within the same single pane of glass experience

12

Cisco Nexus Dashboard
Release Notes

This document. Provides release information for the Cisco Nexus Dashboard product.

Cisco Nexus Dashboard
Hardware Setup Guide

Provides information on physical server specifications and installation.

Cisco Nexus Dashboard
Deployment Guide

Provides information on Cisco Nexus Dashboard software deployment.

Cisco Nexus Dashboard
User Guide

Describes how to use Cisco Nexus Dashboard.

Cisco Nexus Dashboard and Services APIs

API reference for the Nexus Dashboard and services.