Description

Users may see the "VMHost RE x Disk y Label Missing" minor alarm on a device running Junos OS when the labels on the disk, which refer to the set from which the Virtual Machine (VM) booted up, in the Routing Engine (RE) are missing.

This article explains that you can use the "echo" command to bring back the labels and clear the alarm.

Symptoms

The Routing Engine has two sets of disks, which are named SDa (Disk-1) and SDb (Disk-2). You can verify these disks by using the show chassis hardware details or the show vmhost hardware routing-engine both commands.

labroot@router-re0> show vmhost hardware routing-engine both 
re0:
--------------------------------------------------------------------------

Compute cluster: router-re-cc

  Compute node: router-re-cn                   
   Hardware inventory:
    
    
    Item       Capacity             Part number               Serial number             Description
    DIMM 0     8192 MB              VL31D1G63F-K0SE-JN        50044903                  DDR3 1600 MHz
    DIMM 1     8192 MB              VL31D1G63F-K0SE-JN        50044133                  DDR3 1600 MHz
    DIMM 2     8192 MB              VL31D1G63F-K0SE-JN        50044150                  DDR3 1600 MHz
    DIMM 3     8192 MB              VL31D1G63F-K0SE-JN        50044149                  DDR3 1600 MHz
    Disk1      50.0  GB             SFSA050GQ1AA2TO           000060153151A1000061      SLIM SATA SSD
    Disk2      50.0  GB             SFSA050GM3AA2TO           000060157867B1000146      SLIM SATA SSD

re1:
--------------------------------------------------------------------------

Compute cluster: router-re-cc

  Compute node: router-re-cn                   
   Hardware inventory:
    
    
    Item       Capacity             Part number               Serial number             Description
    DIMM 0     8192 MB              VL31D1G63F-K0SE-JN        49398543                  DDR3 1600 MHz
    DIMM 1     8192 MB              VL31D1G63F-K0SE-JN        49398503                  DDR3 1600 MHz
    DIMM 2     8192 MB              VL31D1G63F-K0SE-JN        49398526                  DDR3 1600 MHz
    DIMM 3     8192 MB              VL31D1G63F-K0SE-JN        49398517                  DDR3 1600 MHz
    Disk1      50.0  GB             SFSA050GQ1AA2TO           000060153151A1000041      SLIM SATA SSD
    Disk2      50.0  GB             SFSA050GM3AA2TO           000060156705B1000227      SLIM SATA SSD

You may see the following minor alarms on the device when the labels (which refer to the set from which the VM booted up) on the disk in the RE are missing. Note that these alarms might be transient in nature and may be reported for one disk or both disks.

XX@XXXX> show chassis alarms
1 alarms currently active
Alarm time               Class  Description
xxxxx xxxx UTC  Minor  VMHost RE x Disk y Label Missing

x is RE slot
y is Disk slot  (there are two disks on the RE).

In the "VMHost RE x Disk y Label Missing" alarm, "x" is 0 for RE0 and 1 for RE1 and "y" is 1 for disk 1 and 2 for disk 2.

Solution

Use the following command to get back the label for Disk-2, which is SDb.

 
<< Use request system snapshot and/or request vmhost snapshot partition (depending on the RE that is reporting the alarm)
 
start shell user root 
vhclient -s 
lsblk

echo 1 > /sys/class/block/sdb/device/delete
dmsetup remove jvg_S-jlvmrootrw
dmsetup remove jvg_S-jlvmjunos
dmsetup remove jvg_S-jlvmvm
dmsetup remove jvg_S-jlvmspare

lsblk

for host in /sys/class/scsi_host/*; do echo "host -> $host"; echo "- - -" > $host/scan; sleep 1; ls -l /dev/sd*;done
 

Note: Modify the command "echo 1 > /sys/class/block/sdb/device/delete" to "echo 1 > /sys/class/block/sda/device/delete" if Disk-1 (SDa) is reporting the alarm.

Modification History

2022-09-13: Initial publication