VMware Cloud Community
Rejaine
Contributor
Contributor
Jump to solution

vCENTER appliance restarting and changing configuration from scsi0.present = "true" to "false"

Hello,

I'd like some help.

My vCenter is restarting itself and when this happens, the following message appears in the log:

  

              [msg.Backdoor.OsNotFound] No operating system was found. 

              If you have an operating system installation disc, you can insert the disc into the system's CD-ROM drive and restart the virtual machine.

The disk and configuration files are there and whenever this happens, but  the parameter scsi0.present is changed from TRUE to FALSE

I force the vm to turn off, change it back to TRUE, I'll call you again and everything will go back to normal.

The problem occurs  during a backup job of virtual machines (I am using DataProtector/VEPA). Sometimes vcenter restarts during copying, other times during snapshots creation or removal

This started to occur after I tried to change the configuration of the VM to use CBT (Changed Block Tracking, to allow faster differential/incremental backup), but even changing back configuration to ctkEnabled = "FALSE" and scsi0: 0.ctkEnabled = "false" on all virtual machines and disable CBT on backup application,  the problem persists

Attached is the vmware.log file for review

I´m using vCenter/ESXi 5.5 

Thanks any help.

0 Kudos
1 Solution

Accepted Solutions
Rejaine
Contributor
Contributor
Jump to solution

Problem SOLVED.

Nnothing to do with CTB,  ... the problem is that there were some jobs set up to use vcenter as host backup ..

I changed the host backup to another server (a media gateway) and stopped the problem.

Thank you all for the tips.

View solution in original post

0 Kudos
6 Replies
SupreetK
Commander
Commander
Jump to solution

Did you already try creating a new VM and attaching the existing disks?

Cheers,

Supreet

0 Kudos
Rejaine
Contributor
Contributor
Jump to solution

I did not consider doing this, but as it is vcenter, how should I proceed since I can not use the VCenter GUI (I would have to turn off vcenter to make this change, right?)

Is there any way to do this using the ESXi command line?

0 Kudos
SupreetK
Commander
Commander
Jump to solution

Command line would be difficult. Identify the host on which the vCenter VM is running, login to the host directly using the vSphere client and power off the VM. On the same host, create a new VM with the same specs (RAM, vCPU and the type of NIC adapter) and attach the existing disks from the powered off VM in the same order of SCSI node ID. Once the new VM is deemed to be working fine, you can unmount the disks from the old powered off VM.

Cheers,

Supreet

0 Kudos
Rejaine
Contributor
Contributor
Jump to solution

Well, I'll try to do that.

The strange thing is that when I try to back up only VCENTER, the following error appears after the snapshot is created (during reconfiguration virtual machine task):

     Invalid configuration for device '0'

     Cannot remove virtual disk from the virtual machine because it or one of its parent disks is part of a snapshot of the virtual machine.

There was no other snapshot before the start of the backup (orphan) and I did not understand why the backup procedure wants to remove disk 0 from VCENTER ..

Maybe this is what is causing VCENTER to be restarted during some other backup job and getting the disk scsi0: 0.present = FALSE

0 Kudos
Rejaine
Contributor
Contributor
Jump to solution

It's very strange. When I try to back up any other virtual machine, vcenter loses disk 0 ...

the following message appears in the wmware.log of the vcenter:

2018-09-04T17:14:02.150Z| vcpu-1| I120: CDROM: Emulate GET CONFIGURATION RT 2 starting feature 65
2018-09-04T17:14:02.150Z| vcpu-1| I120: CDROM: Emulate GET CONFIGURATION RT 2 starting feature 65
2018-09-04T17:14:02.151Z| vcpu-1| I120: CDROM: Emulate GET CONFIGURATION RT 2 starting feature 42
2018-09-04T17:14:02.151Z| vcpu-1| I120: CDROM: Emulate GET CONFIGURATION RT 2 starting feature 42
2018-09-04T17:27:24.375Z| vcpu-0| I120: LSI:Event notification sent for SAS device scsi0:0...
2018-09-04T17:27:24.390Z| vcpu-0| I120: Destroying virtual dev for scsi0:0 vscsi=21996
2018-09-04T17:27:24.390Z| vcpu-0| I120: VMMon_VSCSIStopVports: No such target on adapter
2018-09-04T17:27:24.485Z| vcpu-0| I120: scsi0:0: numIOs = 0 numMergedIOs = 0 numSplitIOs = 0 ( 0.0%)
2018-09-04T17:27:24.485Z| vcpu-0| I120: Closing disk scsi0:0
2018-09-04T17:27:24.491Z| vcpu-0| I120: DISKLIB-VMFS  : "/vmfs/volumes/554cb83b-2add89f1-d3c3-f0921c0108b0/VM_VCENTER01/VM_VCENTER01-flat.vmdk" : closed.

I found a similar report using DataProtector and says that this occurs when using something related to hott add disk, but I'm not using it ...

The virtual machine loses its drive. - Micro Focus Community

I had tried using CTB previously but due to the problems I disable CBT for all vm's via PowerCLI ... And do this to make sure that there was no machine with the CBT enabled:

      $ vms = get-vm | ? {$ _. ExtensionData.Config.ChangeTrackingEnabled -eq $ true}

      echo $ vms

The DataProtector is now also with default settings (WITHOUT CTB) for all backup jobs, but vcenter is still losing disk 0 during backup.

I already deleted the backup jobs from the dataprotector, reconfigured all again ...  rebooted ... I do not know what to do ...

0 Kudos
Rejaine
Contributor
Contributor
Jump to solution

Problem SOLVED.

Nnothing to do with CTB,  ... the problem is that there were some jobs set up to use vcenter as host backup ..

I changed the host backup to another server (a media gateway) and stopped the problem.

Thank you all for the tips.

0 Kudos