11-16-2011 03:07 AM
I have backed up a linux virtual machine (without GRT).
While trying to restore this Linux VM with Vmware redirection I noticed that if you select reassigning Network connections, only one network adapter card is restored.
However, doing a restoration without reassigning Networks well works.
I guess this will be taken in account in upcomming versions.
11-17-2011 05:54 AM
Do you have a support case open for this issue? If not, can you get one opened?
This really needs to be investigated further.
Also, what version of Linux are you using here?
12-02-2011 03:02 AM
Hi, yes I'm able to open a case but as this is not 'so' important for me, I just dropped this here so that someone could take this in account for future releases.
Regarding the Linux, uname tell me : SunOS cteraportal 5.11 NexentaOS_134f i86pc i386 i86pc Solaris.
BTW, I noticed recently another issue (I fixed it locally): If a virtual machine has vmdk's corresponding to the maximum file size than the datastore can accept (256Gb in my case), I experienced backup issues as the snapshot can't be created!
I fixed things by migrating all vmdks to a datastore accepting bigger files (eg having a vmware block size of 8MB instead of 1MB).
May be BackupExec could includes some testing so that to warn user by error message other than:
Fin du travail : jeudi 1 décembre 2011 à 16:59:42
Etat de fin : Echec
Erreur finale : 0xe00095a4 - L'opération a échoué car le centre virtuel ou le serveur ESX a signalé un problème lors de la création ou de l'accès à un ou plusieurs fichiers nécessaires.
Catégorie d'erreur finale : Erreurs des ressources
Pour plus d'informations sur cette erreur, reportez-vous au lien V-79-57344-38308
12-02-2011 03:53 AM
The disk space issue is a VMware snapshot requirement not a Backup Exec issue and as such it would be expected that if your datastore is full that a snapshot cannot be created.
12-05-2011 12:20 AM
The issue is that the datastore was not full! the main issue was that (depending the block size the datastore was created) the Virtual machine have a vmdk having the maximum file size the datastore could accept (256Gb in my case)... and despite the datastore has about 1Tb free space, the snapshot can't be taken ... I'm VCP certified and have no such difficulties to fix... but for a non specialist facing BE error message I guess fixing is not obvious.
Hence suggestion to add (in that case) a more comprehensive error message.
12-05-2011 02:48 AM
Ok but is this a VMware snapshot limitation that is documented on VMware's Knowledgebase or something unique to Backup Exec's requirements?
And it if it documented in a VMware KB, then if you can provide the KB article I will start some internal queries about the possibility of a more useful error condition in Backup Exec.
Edit : Ok I found the limitaions as they are specified in
With regards the snapshot do you happen to know if the problem occurs because the snapshot file will be slightly larger than the base vmdk file, or is it because the limitation is on total size of all files relating to the VMDK?
Hmm Edit 2; Snapshot limitations (and how to solve) are provided in a VMware KB (that is referenced in the above document)
As such it is a VMware limitation. I assume that creating a snapshot in the vSphere Client wil also fail in this situation - please confirm this?
12-06-2011 01:11 AM
The issue was initially related to the file size, because it was the maximum allowed, it won't allow taking a snapshot.
Therefore, I migrated that file to a datastore having a bigger block size which allowed me to take snapshot. However, at one time, I had 1 vmdk (the 256GB one) on the old datastore and the virtual machine main folder (with the system disk) on the bigger datastore. At that time, I was able to do manual snapshot using the vsphere GUI while the BE snapshot won't works.
As this is a production environnment, I can spend time to revert so that to confirm. However, I thing that if you create an environment composed with:
1 datastore (8MB block size) > create a VM with system disk on it
1 datastore (1MB block size allowing maximum vmdk file of 256GB) > add to the VM, a 256GB vmdk on that second datastore.
Taking a manual snapshot should works while (as far as I noticed) BE snapshot fail.
12-06-2011 01:31 AM
Ok so all we can do currently is document this - as in reality it is a VMware limitation. Thanks for the details you provided.
EDIT: artcile now available
12-07-2011 12:34 AM
Thank you. I'm always happy to help.