Hi,

In my environment we use Symantec Netbackup 7.1. We backup VM's using our vCenter server to a Netbackup 5000 device & the backups are then copied out to our Disaster Recovery site over a 100MB link.

In our Disaster Recovery site we have 3 ESXi 4.1 hosts & no vCenter server. Is it possible to restore a VM directly to a host instead of going through vCenter?  

Thanks. 

As title says, we have a server that had 146Gb mirrored SAS drives and we used Acronis Backup and recovery Advanced server to save a disk image from the raided mirror. After replacing the drives with the new ones and recreating the mirror, recovery of the disk image back on the same server generates this error. (We did use Disk and not volume imaging. ) No bios settings changes or anything of the sort.

Bank6: Not a VMware Boot bank

No hypervisor found.

The original install of the VMware esx5i was to the raid drives and not usb.

What can be done to "Activate or recover" this server and get it back online?

Thanks

Hi Everyone,

I'm currently evaluating backup options for our VMware estate, and diving into the options is causing a few headaches.

Bit of background info - about 10 ESX 4.1 hosts, VCenter Server 4.1, 100+ VM's, Fibre Channel SAN Storage.

I've found the following options:

VMware Data Recovery - As far as I know, this has been superseded by VMware Data Protection?

VMware Data Protection - Similar to VMWare Data Recovery in terms of technology and operation.

VMware Consolidated Backups - After reading several contradicting articles, I'm struggling to see where this fits in. Some people say it VCB doesn't *actually* back up anything.

So, I think based on my environment and version of vCenter and ESX, my primary choices will be:

VMware Data Protection

Veeam

Can someone briefly summarise the differences with the three listed products and how they fit in with a Virtualised backup strategy please?

Thanks,

David

Hello to all Community Users (this is my first post),

We are small company and since our computing needs recently has growed we decided to virtualize our infrastructure. Our plan is to deploy one (for start) VSphere 5.1 host (2x 6C CPUs, 32GB+ RAM, 4.5TB internal SAS HDD space in RAID) as base for 3-4 production VMs. As a backup solution I planned to use new VSphere Data Protection (VDP) feature in VSphere 5.1 Essentials Plus but deeper analysis revealed some disadventages of VDP (fixed-size datastore, file level restoration does not support EXT4 partitions and so on). It made me think about another backup solution - currently I'm concerning Veeam Backup & Replication 6.5. As a part of backup solution I'm planning to deploy another physical machine - 6.0 TB in RAID NAS based on FreeNAS/Openfiler with NFS share mounted as datastore in VSphere 5.1 host. On this NFS datastore I'm planning deploying Win2008/2012 VM with installed Veeam Backup & Replicaction (backup server) with 2 VMDKs: one smaller for Win2008/2012 and one huge for Veeam backup files (thin provisioned 5.5 TB NTFS VMDK). My qestion is if this planned solution is optimal and safe? Currently we have no budget to buy specialized SAN/NAS appliance, that's a reason I'm planning using FreeNAS/Openfiler with NFS. In the event of VSphere host failure I'm still having all VMs backed up on VBR backup server VM on NAS. After host is brought up I should be able to add VBR VM to vCenter Server and recover production VMs to VSphere host's internal datastore.

PS. Downtime is accepted during Disaster Recovery/host failure.

Thanks in advance for comments and suggestions.

I'll say sorry up front if this topic has been talked about to death but i need/want to ask for any best practices and include my current setup

I will preface to say i work in a public library and we are on a budget.

Currently i have 3 ESXi boxes on ThinkServer T440 boxes 2 production one backup. All the same gear, 500GB internal drive for ESXi and 32GB ram.

I am using a netgear ReadyNAS 2200 8TB over 4 drives unit... and presently all is well, and running nice.

My concern is what to do if the readynas fails i currently am exploring many options but wanted to ask if anyone had any suggestions, maybe close application etc..

I am currently toying with the idea of getting another readynas 2200 with same specs and attempting their backup feature which should bascially mirror the two nas but i am still looking for any research to see how that has worked for other people.

Another option is to take down the vms manually and then copy them to another large store on a weekly basis, i will do this for piece of mind if this is my only option but a more active or automated solution would be preferable.

I am running 2 AD 2008 servers, 1 windows 2008 web server, 5 ubuntu 12.04 servers (backing up core folders), and 2 windows xp machines.

Where i sit is if two drives fail in the Readynas i'm SOL. Otherwise this has been the most wonderful experience as a sysadmin.

Thanks for your time

My setup:

vSphere 5.0.0, 469512

4 VM Instacnes, all being backed up using Veeam 6.1

Using Veeam 6.1 Backup and Replication

Monthly Full with incrementals.

My backup jobs are failing on ONE of the 4 instances.  A cruical one at that.

Veeam support indicated my snapshots are the cause which is not a Veeam support issue but rather VMWare's support issue.  When I go into my vSphere Client, I see no snapshots in the snapshot manager.  However, I saw this error in my events:

Warning message on Email. This virtual
machine has 255 or more redo logs in a
single branch of its snapshot tree. The
maximum supported limit has been
reached, creating new snapshots will not
be allowed. To create new snapshots,
please delete old snapshots or consolidate
the redo logs.
warning

Then I went into the folder in the datastore and saw all the Email-000001.vmdk and Email-000001-ctk.vmdk through Email-000255.vmdk and Email-000255-ctk.vmdk files.  I then noticed the hard drive files the Hard Disk 1, 2 and 3 are using Email-000255.vmdk respectively.

How can I resolve?  I read around Google I two options:

1. consolidate but I don't have enough space nor time for as I understand it could take days.

2. Create a new VM and copy the disks over, but which ones?

3. Use vCenter Converter to convert it to another instance, although it keeps erroring out at Checking Free Space at Destination after 30 minutes.

Can't I just delete all the Email-000001.vmdk and Email-000001-ctk.vmdk files?

I'm configuring an environment which uses Veeam for the backups.

In order to make the snapshots used by the backup tool application consistent we implemented pre-freeze scipts.

We notice however that these have a 15 minute hard-limit in their execution time. Stopping some application servers may take quite some more time to achieve.

Is there a way to alter this time-out setting?

We are small users with only 3 hosts. 2 hosts are running ESXi 4.1, and I am in the process of installing a new third host. We currently rely on Data Recovery and so the new, allegedly more robust, Data Protection appliance was of particular interest. I am therefore trying to upgrade from vSphere 4.1 to 5.1. In order to do so I have to stop using Data Recovery and start using Data Protection instead ( They cannot bee run in parallel as vCenter 5.1 does not support Data Recovery which requires a vCenter connection to work, and vCenter 4.1 does not support ESXi 5.1 ).

I have installed a Data Protection appliance on the new ESXi 5.1 host, but it returns the error " IP/DNS could not be reolved" at the configuration stage.

We are using a typical Windows Server 2003/2008 active directory single domain and the relevant DNS entries have been added manually for all the VMWare Linux appliances. vCenter 5.1 is on a Windows 2008 R2 server. The first problem is that it is not clear whether vDP is case sensitive ( Windows AD is of course not case sensitive, so it would seem crazy for it to be so. However, I have seen suggestions that it is. For now I am assuming that it is not.)

The usual ping and nslookup checks reveal the following:

Pings work between vDP and vCenter for IP, hostname and FQDN.

Ping and nslookup work from Windows servers and PC's for all 3 flavours, whether to a windows machine or Linux appliance.

However, ping and nslookup from a VMWare Linux appliance works for IP and FQDN, but DOES NOT work for a hostname. This is the same from vDP 5.1, vDR 2 or vMA 4.1. Nevertheless vMA and vDR both work fine. It is only vDP which seems to have problems as a result.

I have submitted a VMWare support request a week ago but so far no joy. If I cannot resolve this soon I will have to give up and go back to vSphere 4.1 for our new host.

Is there anybody else with a Windows 2008 AD who has installed Data Protection please?

Is there any way of adding the domain suffix to the hostname automatically in a Linux appliance?

What is the best way to make a full guest OS backup in the case that the hardware failed? I want to be able to simply re-install the hypervisor and restore the guest OS instead of reinstall everything from scratch. Can I do all this through vSphere Client or do I need a third-party software to do this?

When I try to registration a new VDP 5.5 on vCenter 5.5 I have the error: Unable to find this VDP in the vCenter inventory.

For user, I have used "administrator@vsphere.local" or "vsphere.local\administrator" or "mydomain\administrator" and the correct password for all user. I don't understand where is the problem. The vCenter is on the windows 2008R2 64bit and I have deployed the VDP from webclient and vsphere client into different tests.

Please help me!

Thanks!

Francesco

been having some odd happenings the last few days where services would be stopped in the morning on 2 of the VDP appliances, usually management and scheduled

this morning 4 services were stopped on both.

tried reboots on both (took a log bundle from one first)

after more than an hour at the 10 minutes initializing proxies message i logged a call with vmware for help as these are production backup servers.

waited several hours for a return email (after asking for a phone call)  in the meantime reset the machines again.

support not much help, he just suggested it can take a couple of hours to boot and maybe take memory to 8GB for a faster boot.

after about 3 hours the console showed a complete boot, but no web access and no access through vc web client.

shutdown and added memory to both about 3 hours later same thing, console shows completed boot but no access.

logged into the console and checked the disks and found /space partition is 100% full.  another vdp that is still working is about 5% full.

took a full ls of the partition and found a 32GB (the partition is 79GB) called mcs_data_dump.sql in path /space/avamar/var/mc/server_data

does not seem like something that should be there.

can anybody suggest if it is ok to rm this file to get the space back so i can get these backups going again.

Hello!

I have vCenter and ESX 5.1.

I have installed VMware Data Protection 5.5. Then the backup starting i see next error:

An unexpected error occurred with the following error code: 10014. More information may be available in the client logs which can be downloaded from the configuration application

(https://<VDP hostname>:8543/vdp-configure).

error

27.12.2013 13:26:34

VDP: Backup Job

vSphere Data Protection 5.5

What is it mean&

how to optimize backup time, deduplication, performance of appliance.

Thanks in advance

ok this is the situation (ESXi 5 /VSphere client):

VM (WS2008) didn't boot anymore.

I've managed to download via Data Storage Browser the big SERVER-flat.vmdk file plus 14 out of 15 other file sets containing snapshots.

(SERVER-000002.vmdk to SERVER-000015.vmdk + SERVER-000002-delta.vmdk to SERVER-000015-delta.vmdk).

However there was no way to save SERVER-00001.vmdk and the associated delta file and then the server failed completely (data corruption after the failure of one disk in the Raid array, etc.)

I've rebuild the server on another machine from the -flat file (renaming FILENAME inside SERVER.VMX  from SERVER-000015 to SERVER) but this brings me back to the initial server configuration (one year ago).

When I try to add the snapshot files (even renaming the 000002 into 000001  or editing the vmdk file (ParentFileNameHint) it always returns an error.

Error message from esxi502: Reason: The parent virtual disk has been modified since the child was created. The content ID of the parent virtual disk

does not match the corresponding parent content ID in the child.

Cannot open the disk '/vmfs/volumes/51312345-e57af280-etc-etc/SERVER.domain.it/Server-000002.vmdk' or one of the snapshot disks it depends on.

Question 1: Is there any way to incorporate the snapshots, jumping from SERVER-000002 to SERVER.vmdk, seeing that the first one is missing?

Question 2: is there any program that can read/recover data from single  vmdk files? (tried already Systools VMware recovery but this does not work with files saved from VM to local disk as the support explained to me)

Any help would be highly  appreciated as I'm desperate, not very familiar with VMware and the data is crucial for us.

Thanks a lot!

Hi,

If anyone can help me that would be appreciated. I have setup VDP to backup my VMs individually. Previously my backup used to work but recently, I keep getting the error saying " VDP : A backup/restore/validate activity did not start within the alloted time" in the Tasks. Even trying to perform a "Backup Now" doesn't seem to have any effect. I also did a check on the server logs via ssh in /usr/local/avamar/var/vdr/server_logs/ but nothing seems to show me to the right path.

Is there some restriction to the time you can set a backup job to start, even within the backup window? I have two jobs on one VDP. The daily job I set for 5PM (maint. window is 6AM-5PM), and I am trying to set the other (a weekly job) to 6PM, but every time I do that it changes the time on it back to 10PM! VERY frustrating... I had the same problem when trying to set the job for 5PM (the time I REALLY would prefer), but I figured that wasn't working because it just didn't like both jobs being set to the same time.

Also, why is there not a way to set a job to be performed on certain days of the week? For instance, I do not want my daily job to run on Saturdays, but with VDP it's all or nothing with the daily, weekly, and monthly options. The reason I need it to not perform the daily backup job on Saturdays is because I have to back up an SQL DB VM on Saturday, and for some reason backing up SQL VMs seems to take a MUCH longer time than normal ones, especially if other backups are running at the same time.

Which leads me to ANOTHER problem.... My backups simply are not completing in time. It seems to usually only be the SQL ones that run MUCH slower as I mentioned, but occasionally that causes one of the other backups to not complete either. Are you telling me that if a backup cannot be completed within the backup window then it cannot be completed at all?? That's ludicrous. The backup system just simply isn't available during the maintenance window? That effectively makes it useless 50% of the day (if maint. windows set to 12 hours, which it seems like it needs to be in order to complete it's tasks).

Hi!

I have a problem with the backup size. Backup is larger than the used space by the VM OS.

VM is configured Thick with Lazy. I have also VMs with Eager. Problem is the same.

Datastor a SAN Luns, wich are configured as Thin.

ESX is 5.5

CommVault Software is Version 10, SP 10.

What we tested out:

- VM with a size of 30GB (container is 80GB large). Full Backup is 30GB. Result is as expected

- We copy 10 GB to the VM, so its 40GB large. Full Backup is 40GB. Result is as expected

- We deleted the 10GB, so the VM is again 30GB large. Full Backup is still 40GB.

Incremental backups are running fine.

CBT is enabled.

Tried unmap, moved to another datastor, but Backup is still 40GB large.

I found in my logs ### QueryChangedDiskAreasPrivate --- Exception during QueryChangedDiskAreas Von Datei /vmfs/volumes/52fb3a1a-949ff648-08ff-0000c994e052/VSA_TEST01/VSA_TEST01_4-000002.vmdk verursachter Fehler. (dont know why its in german).

If you google QueryChangedDiskAreas you find some vmware KBs, all with CBT Problems.

I also found the information that Windows doesnt gave the space back. You can run SDELTE -c (its a windows tool wich zero out the unused space). But in our case we ware still on 40GB full backup.

So my questions to you:

Do we have a CBT Problem?

Or what else can it be?

What can we do to point out our problem?

What can we else try to get the 10GB back?

Thanks in advance

Br

Jan

Hi Team ,

One of my VM guest Machine's backup job is getting failed again and again with snapshot error .

When I checked logs I can see Event Warning  " VM disks consolidation failed "  where currently there is no snapshot is available  on this machine.

But When I manually run the snapshot by using by using quiesce Method and then again run backup Job from Symantec , Backup runs fine 7

We are using Symantec Netbackup 7.6.1 and My VM Machine is running on ESXi01 , VM version 9 , Guest OS MS Windo 2008 R2

Any Help Would be really appreciate

In preparation for upgrading from ESXi 5.5 to 6.0, the best practices upgrade document strongly suggests that "you backup the host system". This is very sound advice, but how do you actually go about doing the backup?

I can't find any VMware documentation on actually backing up the host system prior to an upgrade. I guess to a Unix guru it might be obvious.

Also, to provide some background, we were using VMware Workstation prior to January and decided to migrate to ESXi (VMware vSphere 5 Essentials), so the V6 upgrade is the first upgrade we'll be doing. We have already upgraded VCSA to version 6. It is a "tiny" installation - 1 active host and 6 VM's.

We have backup solutions in place for the VM's using Veeam, so that isn't the issue.

I know with some operating systems (like OpenVMS) we boot from an OS DVD and run what is called a standalone backup to make an image of the system disk prior to upgrades - the concepts are not foreign, just the mechanics with ESXi. I guess I'm expecting we'd do something similar but maybe it isn't that complex.

Thanks for any advice.

Hi,

I need a reality check of my VDP setup.

SITE 1: VDP (A)

SITE 2: VDP (B)

I make primary backups to VDP(A). Then I use VDP replication to VDP (B) that is placed on a separate site.

If site 1 and VDP (A) crashes or is otherwise destroyed, how do I recover the replicated backups from VDP (B) and how do I test it?

Regards Marcus