After upgrading an ACX7348 device to Junos OS Evolved 25.4R1 or later, the chassis redundancy configuration statement failover on-loss-of-keepalives generates a warning message "statement ignored: unsupported platform (acx7348)" during commit. This occurs because chassis redundancy failover options including on-loss-of-keepalives, disk-read-threshold, disk-write-threshold, not-on-disk-underperform, and on-disk-failure are no longer supported on ACX7348, ACX7332, and ACX7059 platforms running Junos OS Evolved. This change is by design and does not negatively impact redundancy functionality.

When committing the chassis redundancy configuration on ACX7348 running Junos OS Evolved 25.4R1 or later, the following warning appears:

chassis {
    redundancy {                        
        routing-engine 0 master;
        routing-engine 1 backup;
        ##
        ## Warning: statement ignored: unsupported platform (acx7348)
        ##
        failover on-loss-of-keepalives;
        graceful-switchover;
    }
}

The configuration that previously worked without warnings on earlier versions now shows the platform incompatibility warning. The system continues to operate normally with RE redundancy intact.

Starting with Junos OS Evolved 25.4R1, the failover on-loss-of-keepalives statement and other chassis redundancy failover options are not supported on ACX7348, ACX7332, and ACX7059 platforms. This is documented in the official release notes.

No Negative Impact: There is no functional implication from removing this command, as it is not needed per the Junos OS Evolved architecture design:

1. Application Monitoring: Critical applications such as packetio and ppman are monitored by the redundancyd application, which automatically triggers a graceful switchover if these applications fail. The switchover involves a "prepare phase" from the Evo infrastructure.
2. Distributed State Model: Junos OS Evolved uses a Distributed Data Store (DDS) infrastructure where applications publish or subscribe to state objects. This ensures state information is preserved across process restarts, unlike traditional Junos OS where state is stored internally within processes.
3. Switchover Conditions: In Junos OS Evolved, switchover is only allowed when:

• Configuration and DDS are synced between master and backup
• Backup RE is online
• No failed applications exist on the backup RE

Mastership Handling: RE-to-RE communication and mastership switchover are independent in Junos OS Evolved. The RECPLD generates mastership interrupts serviced by the LPC driver as udev events, which mstrd monitors and provides to applications like HwdRE and evo-pfemand.

Recommended Actions:

1. Remove the unsupported failover on-loss-of-keepalives statement from the configuration to eliminate the warning message
2. Retain the graceful-switchover statement, which remains supported
3. Verify RE redundancy status using show chassis routing-engine and show system switchover
4. The redundancy functionality continues to work as designed through the Junos OS Evolved infrastructure

Configuration Example:

chassis {
    redundancy {
        routing-engine 0 master;
        routing-engine 1 backup;
        graceful-switchover;
    }
}

Reference: For additional details, refer to the Junos OS Evolved 25.4R1 Release Notes under "What Changed" for ACX platforms.

https://www.juniper.net/documentation/us/en/software/junos/release-notes/25.4/junos-release-notes-25.4r1/topics/what-changed/acx-what-change-cover.html