Upgrade NetQ Virtual Machines

This page describes how to upgrade your NetQ virtual machines. Note that the upgrade instructions vary depending on NetQ version you’re currently running.

For deployments running:

During the upgrade process, NetQ will be temporarily unavailable.

Before You Upgrade

  1. Verify that Kubernetes is running and the admin app is up:
cumulus@masternode:~$ /home/cumulus# kubectl get pods|grep admin
    netq-app-admin-masternode                            1/1     Running            0               15m

If the output of this command displays errors or returns an empty response, you will not be able to upgrade NetQ. Try waiting and then re-run the command. If after several attempts the command continues to fail, reset the NetQ server with netq bootstrap reset keep-db and perform a fresh installation of the tarball with the appropriate netq install command for your deployment type.

  1. Back up your NetQ data. This is an optional step for on-premises deployments. NVIDIA automatically creates backups for NetQ cloud deployments.

Update NetQ Debian Packages

  1. Update /etc/apt/sources.list.d/cumulus-netq.list to netq-4.9:

    cat /etc/apt/sources.list.d/cumulus-netq.list
    deb [arch=amd64] https://apps3.cumulusnetworks.com/repos/deb focal netq-4.9
    
  2. Update the NetQ debian packages. In cluster deployments, update the packages on the master and all worker nodes.

    cumulus@<hostname>:~$ sudo apt-get update
    Get:1 https://apps3.cumulusnetworks.com/repos/deb focal InRelease [13.8 kB]
    Get:2 https://apps3.cumulusnetworks.com/repos/deb focal/netq-4.9 amd64 Packages [758 B]
    Hit:3 http://archive.ubuntu.com/ubuntu focal InRelease
    Get:4 http://security.ubuntu.com/ubuntu focal-security InRelease [88.7 kB]
    Get:5 http://archive.ubuntu.com/ubuntu focal-updates InRelease [88.7 kB]
    ...
    Reading package lists... Done
    
    cumulus@<hostname>:~$ sudo apt-get install -y netq-agent netq-apps
    Reading package lists... Done
    Building dependency tree
    Reading state information... Done
    ...
    The following NEW packages will be installed:
    netq-agent netq-apps
    ...
    Fetched 39.8 MB in 3s (13.5 MB/s)
    ...
    Unpacking netq-agent (4.9.0-ub20.04u45~1710409093.ea9740d7c) ...
    ...
    Unpacking netq-apps (4.9.0-ub20.04u45~1710409093.ea9740d7c) ...
    Setting up netq-apps (4.9.0-ub20.04u45~1710409093.ea9740d7c) ...
    Setting up netq-agent (4.9.0-ub20.04u45~1710409093.ea9740d7c) ...
    Processing triggers for rsyslog (8.32.0-1ubuntu4) ...
    Processing triggers for man-db (2.8.3-2ubuntu0.1) ...
    

Download the Upgrade Software

  1. Download the upgrade tarball.

    1. On the NVIDIA Application Hub, log in to your account.
    2. Select NVIDIA Licensing Portal.
    3. Select Software Downloads from the menu.
    4. Click Product Family and select NetQ.
    5. Select the relevant software for your hypervisor:
      If you are upgrading NetQ Platform software for a NetQ on-premises VM, select NetQ SW 4.9 Appliance to download the NetQ-4.9.0.tgz file. If you are upgrading NetQ software for a NetQ cloud VM, select NetQ SW 4.9 Appliance Cloud to download the NetQ-4.9.0-opta.tgz file.
    6. If prompted, read the license agreement and proceed with the download.

    For enterprise customers, if you do not see a link to the NVIDIA Licensing Portal on the NVIDIA Application Hub, contact NVIDIA support.


    For NVIDIA employees, download NetQ directly from the NVIDIA Licensing Portal.

  2. Copy the tarball to the /mnt/installables/ directory on your NetQ VM.

Run the Upgrade

Perform the following steps using the cumulus user account.

Pre-installation Checks

Verify the following items before upgrading NetQ.

  1. Confirm your VM is configured with 16 vCPUs. If your VM is configured with fewer than 16 vCPUs, power off your VM, reconfigure your hypervisor to allocate 16 vCPUs, then power the VM on before proceeding. For cluster deployments, verify these requirements on each node in the cluster.

  2. Check if there is sufficient disk space:

cumulus@<hostname>:~$ df -h /
Filesystem      Size  Used Avail Use% Mounted on
/dev/sda1       248G   70G  179G  28% /
cumulus@netq-appliance:~$

NVIDIA recommends proceeding with the installation only if the Use% is less than 70%. You can delete previous software tarballs in the /mnt/installables/ directory to regain some space. If you cannot decrease disk usage to under 70%, contact the NVIDIA support team.

  1. Confirm that the NetQ CLI is properly configured. The netq show agents command should complete successfully and display agent status.

  2. Ensure that the required ports are open according to your deployment model.

If you are upgrading a cluster deployment to NetQ 4.9.0, you must open TCP port 36443 for Kubernetes control plane operations.

  1. Run the following command to ensure that the NetQ version is consistent between the netq-agent package, the netq-apps package, and the tarball you downloaded in the previous section:
cumulus@<hostname>:~$ dpkg -l | grep netq
ii  netq-agent                      4.9.0-ub20.04u45~1710409093.ea9740d7c amd64        Cumulus NetQ Telemetry Agent for Ubuntu
ii  netq-apps                       4.9.0-ub20.04u45~1710409093.ea9740d7c amd64        Cumulus NetQ Fabric Validation Application for Ubuntu

Upgrade Using the NetQ CLI

  1. Run the appropriate commands for your deployment type:
cumulus@<hostname>:~$ netq upgrade bundle /mnt/installables/NetQ-4.9.0.tgz

If this step fails for any reason, run the netq bootstrap reset keep-db command and perform a fresh installation of the tarball with the netq install standalone full command.

Run the netq upgrade command, specifying the current version’s tarball and your cluster’s virtual IP address. The virtual IP address must be allocated from the same subnet used for your master and worker nodes.

cumulus@<hostname>:~$ netq upgrade bundle /mnt/installables/NetQ-4.9.0.tgz cluster-vip <vip-ip>

If you are upgrading from a NetQ 4.8 high availability, on-premises cluster with a virtual IP address, you do not need to include the cluster-vip option in the upgrade command. Specifying a virtual IP address that is different from the virtual IP address used during the installation process will cause the upgrade to fail.

If this step fails for any reason, run the netq bootstrap reset keep-db command and perform a fresh installation of the tarball with the netq install cluster full command.

cumulus@<hostname>:~$ netq upgrade bundle /mnt/installables/NetQ-4.9.0-opta.tgz

If this step fails for any reason, run the netq bootstrap reset keep-db command and perform a fresh installation of the tarball with the netq install opta standalone full command.

Run the netq upgrade command, specifying the current version’s tarball and your cluster’s virtual IP address. The virtual IP address must be allocated from the same subnet used for your master and worker nodes.

cumulus@<hostname>:~$ netq upgrade bundle /mnt/installables/NetQ-4.9.0-opta.tgz cluster-vip <vip-ip>

If this step fails for any reason, run the netq bootstrap reset keep-db command and perform a fresh installation of the tarball with the netq install opta cluster full command.

  1. Confirm the upgrade was successful:
```
cumulus@<hostname>:~$ cat /etc/app-release
BOOTSTRAP_VERSION=4.9.0
APPLIANCE_MANIFEST_HASH=a9d82e8df46178c9a0b3ac17678d4ae8aeba54a89c502fc8042de1f784fc3ef2
APPLIANCE_VERSION=4.9.0
APPLIANCE_NAME=NetQ On-premises Appliance
```
```
cumulus@<hostname>:~$ cat /etc/app-release
BOOTSTRAP_VERSION=4.9.0
APPLIANCE_MANIFEST_HASH=c743bca6bb7ca28a17e7b27559bb13f2098e4d7a810b658bfd248a46fd0e09c5
APPLIANCE_VERSION=4.9.0
APPLIANCE_NAME=NetQ Cloud Appliance
```

Next Steps