Basic VEEAM NAS Backup - Additional Options for Recovery

[bubbawny69]

Hi Friends:

We are a very small business with (2) Dell ESXi hosts, each running their own DAS arrays, VMWARE ESXI 6.7U1.

We have VEEAM Backup & Replication 9.5 Update 4 that is running on the second Dell HOST as a VM (Windows 2016 std with VEEAM B&R installed). This VM running VBR 9.5 backups to a repository on an iSCSI SAN volume located on an external SAN device located in our data center. We have (3) Jobs running backups of (3) VMs, daily incremental, weekly full. Again, we are a very small business with simple infrastructure at this point.

For several years now, we've been running really without any issues, outside of many flaws in our backup architecture driven by lack of financial resources and understanding by the small business president/CEO about the risks to his business, given lack of investment into more adequate backup, archival, etc.

While our VBR daily jobs have been running without issue, as IT Manager, I'm growing concerned about our ability to recover in a crisis situation.

My thoughts were to add an additional SAN device, located internally over our Cisco LAN, but in the back of our 100,000sq.ft' warehouse (different area). My goal is to make available a redundant copy of the backup files so as to be able to recover.

My concern is that even in doing this, if my ESXi host goes down that is running the VBR VM, I'm wondering how I would be able to quickly recover my VBR VM/Backup ESXi HOST from this additional backup copy SAN device.

My other idea was to some how create a job that would EXPORT my (3) VMs to this additional SAN, ready to be imported on to a NEW ESXi HOST in the event of a disaster. (If one or both hosts are faulted, buy new hardware, install ESXi, import VMs, backup and running).

Our simple business is not truly a mission critical operation. And, budgetarily, our CEO/President has been averse to spending on IT infrastructure, no matter the argument. (They've over expanded in the market and are cash flow poor). While I've talked about how a critical disaster could put him out of business, he's just averse to spending any amount of $ further on IT infrastructure, with the exception perhaps of one additional SMB SAN.

So if we buy and add one more SMB SAN, geo-located on the other side of our warehouse, that would at least give us one more backup recovery point. But I'd desire at a minimum to not just have a backup copy, but a fully ready exported VM of at least the VBR VM on that device. So if we had a failure, get new hardware, quickly install ESXi, copy the VM to the datastore, boot, & restore other (2) VMs.

I was hoping for some advice about how to more effectively expand our single disk based backup into something that is a bit more adequate for backup redundancy (onsite) and for more rapid recovery. Additionally, it's also more about archival too...how can we get to a point where I can start saving point in time backups of the VMs more permanently. Today, given my SAN space limits, I can save (14) restore points.

I thank you for your advice. I have been having to work with limited IT $ and do have a basic backup mechanism. But as IT Manager, I worry about the failure of our solo-SAN device and no other backup respository. At least adding another SAN, geo-located elsewhere weould be an improvement. I'd like to get to OFFSITE & archival. Thought of an Amazon S3 bucket, but our Internet PIPE is so tiny, it'd take forever for a byte level backup to an S3-bucket, and I think the S3 costs for about (12TB) of VMDK files would be unaffordable.

Getting another SMB SAN is within our investment reach (I hope!), and I know it'd help with at least having another backup copy elsewhere.

The final question them comes down to whether I'm prepared to recover. And I'd like to make sure that I can...that I can recover to another new ESXi HOST, get my VBR VM up, and recover the other VMs.

If finances weren't a constraint, I'd hire someone to come in and help ensure we have all the right resources, designs, and preparedness. But with a paucity of IT $, I'm simply trying to evolve my next step design for backups into something better than what I have now.

Thanks.

PS. I'm sure I'll get flamed for such a poor mans design. But it is what it is, and I'm only trying to evolve it to help protect my company, given the constraints on IT Spending. Thank you again.

[bdufour]

if the building burns down, or gets robbed, everything is gone. i would focus on getting backups/replicas offsite... even if for now, you had a couple of recent full backups (encrypted) on an external HD or NAS device that could be taken offsite after the jobs are done (not perfect, but given the lack of budget and size it may be a good start, and would give you and the CEO peace of mind)

[bubbawny69]

Yes.

I'm looking to do that as well. I know it's not sexy, but manually having point in time copies of the VMs and backup files on a portable HD and taken offsite is something I intend to do.

Honestly, only once before have I ever met such a penny-pinching and stubborn company. Once before I was in this same pickle. The company failed to do key backups of a system, had a disaster that cost them HUNDREDS of THOUSANDS. Only then did they realize it was their mistake to not listen to the myriad of prior proposals to perform adequate investment into DR and archival. (It was in a sad way a good thing for the IT Team b/c they finally approved significant investment to protect their other systems).

Anyways, my rant isn't so much about the penny pinching as much as it is the HOW or WHAT WOULD BE RECOMMENDED in terms of VBR jobs to do this.

I think minimally, to be prepared for DR, I need to have a copy of the backup repository. I need to have a copy of my BACKUP VM, ready to be used with the backup repository to recover the other systems. I need to then figure out how I can use VBR or VMWARE scripts to routinely ensure that both the backup repository is stored on the second SAN, and that someone regularly scheduled jobs to EXPORT my VBR VM to the secondary SAN is also done.

Is there an option in VBR to schedule a restore job to run periodically that would EXPORT the VBR VM?

If not, does anyone know of a script that would EXPORT my VBR VM as a self-contained, run-ready VM on a storage device (say a network share point or USB Device)?

I think I'd at least sleep right knowing I would at least have:

* a second source of the VBR Repository should the PRIMARY SAN have a disaster

* a run-ready standby copy of the VBR VM that could be quickly be made operational on a new ESXi host. In conjunction with this, would be able to utilize the second source VBR repository to recover the remainder of the operating environment.

Ideally, my plan is to come up with a mechanism here, to implement it, and then to TEST recovery to ensure that I am able to utilize these tools to recover from failure in a DR situation.

[bdufour]

if youre going to do a secondary SAN, you can simply replicate your VMs to that SAN - that way if something happens you can failover to the replica VMs instead of waiting to restore a backup. additionally, i would have backups as well, stored where you feel comfortable. but in a true disaster situation, id rely on a ready to run replica vs restoring a backup.

you can also use veeams surebackup platform to test your backups and replicas - works well.

[foggy]

Is there an option in VBR to schedule a restore job to run periodically that would EXPORT the VBR VM?

You'd better use configuration backup and restore for that purpose.