You can install Commvault File System (CVFS) and operating system (OS) updates on Commvault Grid nodes to keep the environment up to date with the latest VaultOS or Red Hat updates, depending on your deployment model. Updating CVFS or OS also updates the latest Commvault Maintenance Release on the nodes.
OS updates are installed from the Command Center. The process includes:
-
VaultOS or Red Hat updates
-
Commvault File System (CVFS) RPM updates
The process does not update any third-party RPMs that were manually installed on the nodes.
Update Modes
| Upgrade Mode | Description | Approximate Duration | Impact |
|---|---|---|---|
| CVFS Disruptive Mode (Quick Upgrade) | Installs updates on all nodes simultaneously. | 90–120 minutes (total) | Services may be temporarily unavailable during the upgrade. |
| CVFS Non-Disruptive Mode (Rolling Upgrade) | Installs updates one node at a time. Backup and restore operations continue on the remaining nodes during the upgrade. | Longer than a Quick Upgrade (depends on cluster size) | Minimal disruption. Backup and restore operations continue on the remaining nodes. |
| OS Disruptive Upgrade (DU) | Optimizes the upgrade by performing most operations in parallel. Kernel RPMs are upgraded and nodes are rebooted one at a time to maintain cluster availability if a failure occurs during the upgrade. | Approximately 45 minutes per node | All cluster jobs are placed in a pending state, and no new jobs are initiated because the cluster and Commvault Grid MediaAgents are placed in maintenance mode during the upgrade. |
| OS Non-Disruptive Upgrade (NDU) | Performs the upgrade sequentially, one node at a time. Only the node being upgraded is taken offline, allowing the remaining nodes in the cluster to continue servicing backup and restore operations. | Approximately 60–75 minutes per node | Backup and restore jobs running on the node being upgraded may enter a pending state. The jobs automatically resume after the upgrade completes and services restart on the node. |
By default, CVFS updates are installed using Disruptive Mode (Quick Upgrade).
Note
-
A non-disruptive upgrade maintains overall cluster availability, but it does not guarantee uninterrupted job execution.
During the upgrade, jobs running on or assigned to the node being upgraded can be temporarily interrupted and automatically retried. You might see messages similar to Infrastructure software update in progress. Job will be retried. These retry events are expected behavior, and jobs typically complete successfully after the upgraded node becomes available or another node resumes processing.
If you use custom monitoring, alerting, or workflow automation that reacts to temporary job failures or retry messages, consider excluding these expected events during the upgrade window.
-
To install OS updates from the Command Center, the CommServe server must be outside the Commvault Grid cluster.
Before You Begin
-
Ensure the CommServe and all Commvault Grid nodes are running Commvault Platform Release 2024E (11.36) or later, with the latest maintenance release. Both must be on the same release.
-
Verify that a smart server group containing all Commvault Grid nodes is available.
Note
-
When you upgrade the CommServe to Commvault Platform Release 2024E (11.36) or later, a smart server group is automatically created for the Commvault Grid storage pool.
-
Before starting a non-disruptive upgrade (NDU), verify that the disk usage on all nodes in the cluster is below 80%. If disk usage exceeds this threshold, the upgrade will fail.
-
For new installations, the smart server group is created during storage pool configuration.
-
If a group is not available, create one with the rule type HyperScale nodes of the storage pool. For details, see Adding a Smart Server Group by Using Automatic Association.
-
Verify that Commvault services and CVFS services are running:
-
# commvault list- Confirm that all services are running. -
# /usr/local/hedvig/scripts/showmembers.exp- Confirm that the output showsUNREACHABLE MEMBERS:0.
-
-
If using a tape library, update the Linux tape driver (lin_tape) when the OS kernel is upgraded.
For details, see Tape Library Support for HyperScale X.
-
Enable root access on all nodes.
For instructions, see Restricting Root Access on HyperScale X.
-
If your Commvault Grid nodes use Fibre Channel (FC) connections, run the following command and resolve any issues before proceeding with the upgrade.
multipath -F
Procedure
Step 1: Verify Remote Cache Configuration
By default, a remote cache is created on one node when a storage pool is configured. Verify that it is available.
-
From the Command Center navigation pane, go to Manage > Maintenance.
The Maintenance page appears.
-
Click Download/copy software.
The Download/copy software section appears.
-
Click Schedule Job.
-
In Remote caches to sync, expand Servers and verify that one of the nodes in the cluster is listed.
Note
If there are multiple Commvault Grid clusters on a CommServe, each cluster requires a separate remote cache.
If no node is listed, set up a remote cache on one of the nodes. See Setting Up a Remote Cache on a Client Computer.
-
Click Cancel.
Step 2: Download Updates and Populate the Remote Cache
-
From the Command Center navigation pane, go to Manage > Maintenance.
The Maintenance page appears.
-
Click Download or copy software.
-
Click Run job
. -
Select Download software using Internet or Local Copy.
-
Under Download installation media, select the required feature release or maintenance release.
-
In Operating system options, select Unix, and then select Linux x86_64.
Note
The latest OS updates are always downloaded, regardless of the selected option.
-
In Remote caches to sync, select the Commvault Grid node with the configured remote cache.
-
Click Download.
Monitor the download job to ensure that the download and synchronization complete successfully. For details, see Viewing Jobs.
Step 3: Install OS Updates
-
From the Command Center navigation pane, go to Manage > Infrastructure > Server groups.
The Server groups page appears.
-
For the smart server group containing the Commvault Grid nodes, click the action button
, and then select Upgrade software. -
In the Confirm software upgrade dialog box:
-
Select Install operating system updates.
-
(Optional) Select Install storage updates to install CVFS binaries with the OS updates.
-
(Optional) Select Rolling upgrade (non-disruptive) to upgrade one node at a time while maintaining overall cluster availability.
-
Backup and restore operations remain supported throughout the upgrade. Backup and restore jobs running on the node being upgraded may enter a pending state. The jobs automatically resume after the upgrade completes and services restart on the node.
-
Jobs using the node that is currently being upgraded can be temporarily interrupted and automatically retried.
-
You might see messages such as Infrastructure software update in progress. Job will be retried. These messages are expected during the upgrade.
-
If you install only OS updates, the installation is always disruptive.
-
-
-
Click Yes.
Monitor the upgrade job to confirm that the installation completes successfully. For details, see Viewing Jobs.
Note
During installation:
-
Updates are applied to one node at a time.
-
Each node is temporarily taken out of service while updates are installed and the node is rebooted.
-
Jobs that are using the node being upgraded can be temporarily interrupted and automatically retried. These retries are expected and typically do not require user intervention.
-
Depending on workload timing, new jobs can be assigned to a node shortly before it enters the upgrade process and might also be retried.
-
The node hosting the remote cache is rebooted last. After it restarts, the upgrade automatically resumes. Post-upgrade tasks on this node can take additional time to complete.
Expected Behavior During a Rolling Upgrade
During a rolling (non-disruptive) upgrade, the cluster remains available while individual nodes are upgraded one at a time.
The following behaviors are expected:
-
Backup and restore operations remain supported throughout the upgrade.
-
Jobs that are using a node being upgraded can be temporarily interrupted and automatically retried.
-
Job history can include messages such as Infrastructure software update in progress. Job will be retried.
-
Jobs might remain in a retry state for a short period before automatically continuing.
-
Temporary retry messages can trigger external monitoring or workflow automation that interprets retries as failures. If applicable, plan maintenance windows or adjust monitoring rules before starting the upgrade.
In most environments, these retry events are transient, and jobs complete successfully after the upgrade progresses.
Result
-
OS updates are installed on all Commvault Grid nodes.
-
Jobs that were temporarily interrupted during the upgrade are automatically retried and typically complete successfully.
-
To verify that the storage and operating system (OS) updates completed successfully, perform the following steps:
-
From the Command Center navigation pane, go to Manage > Storage > HyperScale X.
The HyperScale X page appears.
- Click the storage pool.
The storage pool page appears.
- Click the Nodes tab.
The nodes in the storage pool are displayed.
-
For each node, click the node name and verify the following information in the Server Information section:
-
CVFS Version: If you installed storage updates, verify that all nodes display the CVFS version of the latest platform release. The version should be the same on all nodes.
-
Kernel Version: If you installed operating system updates, verify that all nodes display the kernel version of the latest platform release. The version should be the same on all nodes.
What to Do Next
-
Disable root access on all nodes. For instructions, see Restricting Root Access on HyperScale X.
-
If the nodes are running RHEL 7.9, update or reinstall the megaraid_sas and smartpqi drivers after applying OS updates. These drivers are kernel modules and must match the running kernel version.