Hourly Snap Shot on a vm recommendation?

[iamxCPx]

Hi guys,
Is it possible to use Veeam to do hourly snapshot/backup of our file server only without impacting performance so that way if I need to restore some files I don't have to go back to 24 hours prior?
I have enabled Windows Shadow Copy on the file server but I don't trust them if we ever got infected by malware (Crypto locker specifically) because it will silently delete those snapshot.
Our file server size is almost 500 Gb so that's why I'm a bit reluctant to do a vm backup every hour because it will take a long time.
I'm open to any suggestions.

I did open a case also with support just in case.
The case #00939434

Thank you all.

[Shestakov]

Hi,
Yes, you can backup your VMs hourly. Some our clients practice it successfully. If backup job duration is less than 1 hour, you are fine.
For the fastest RTO in case of restore, it`s also recommended to use replication jobs.
If you define your goal, I can give you more advice.
Thanks!

[michaelryancook]

Even though you have 500 GB of total data, Veeam would only be backing up the changes from that hour. Do you have any idea what your change rate is? For simplicity's sake you could take your daily change amount divided by the number of business hours to get an average hourly change rate during business hours. Then use that to estimate how long the backup would take. If you had a compatible SAN you could use the SAN integration to ensure the VM snapshot was minimized.

[iamxCPx]

Thanks for the respond guys.

My goal is pretty simple.
I just want to be able to do file level restore to an hour before the disaster struck the file server for now. I'd love to do the whole infrastructure but I don't have the resources to hold all of the data since we are a small business type. You know how that goes with limited budget.

I do have the Backup Copy To jobs setup. One will keep archive weekly for 4 weeks and the other one will keep archive monthly for 4 months.
Is that what you meant by replication? If not, does replication will compress the file or I will need to provide at least 500 GB somewhere? USB drive maybe?

I will look into the change rate number tomorrow since it's already a half day today.
We use Tegile Array for our SAN and it does have a snapshot capability and it's enabled by default. I'll check with them to see if it can be integrated with Veeam.

Thanks!

[Shestakov]

Replication is having another instance of VM on a DR site ready to become a production one.
In your case, if you don`t have an additional host, you can run hourly backups with backup copy job for historical purposes.
For your purposes Forever Forward Incremental Backup seems like the best option.
Thanks!

[iamxCPx]

Backup Copy Job doesn't have the option to do Forever Forward Incremental Backup.
Regular Backup does.

Does that means I have setup another backup job just for the file server to link up with a new backup copy job?
I'm sorry, I'm a little confused here.
I do appreciate the help.


THanks,
Chris

[Shestakov]

Actually by its nature Backup copy job uses forever forward incremental method
But I meant the source job in the previous post.
Anyways, you don`t need an additional job, 2 (backup job and backup copy job) is enough. Thanks!

[iamxCPx]

What I meant by 2 backup jobs was that I have to create a new "hourly" backup job and then a backup copy job to copy that "hourly" back job offsite.

So for example..

My current backup jobs are:
Backup Job #1 - Contains 5 VMs
Backup Job #2 - Contains 6 VMs
Backup Job #3 - Contains 3 VMs

They are running on schedule every night one after the other.

IF I want to create hourly backup job for my File Server then I need to create:
Backup Job #4 - Contains only my File Server VM.
Scheduled it to run hourly.

IF I want to have a copy offsite then I need to setup a Backup Copy Job linked to Backup Job #4 that points to the offsite repository.

Is that sounds about right?

Thank you again all.

[Shestakov]

Yes, sounds good!

[iamxCPx]

Ok, I just experienced a low disk issue on the repository and after investigation, this is what I found.

Backup Job #2 - It includes the File Server as one of the server being backed up, it's setup to use Reverse Incremental with Active Full every Saturday = it will have 2 giant .vbk files.
Backup Job #4 (The Hourly) - setup as Forever Forward Incremental = 1 big vbk (300 Gb) and then 11 small vib (set to backup for 12 hours at hourly rate).

I'm trying to see if there's a better solution than my current setup to save space.
Because as you can see the File Server (300 Gb) is basically being backed up in 3 places in a week (Reverse Incremental, Active Full, and Hourly). 6 places if we want to include the 3 Backup-Copy job that I setup for 3 offsite locations.

Any input would be very appreciated.

Thank you guys.

[Shestakov]

What is the reason to backup the big file server in 2 different jobs(#2 and #4)? I would exclude it from the job using reverse increments.
As for backup copy job, just interested why you need the copy of .vbk in 3 different locations? Usually one is enough.
Thanks!

[iamxCPx]

That's so true. I could just delete the file server from job #2.
My head is still thinking the "Backup Exec" way.
I thought when I set it up to do the hourly for the File Server, it will just use the vbk from job #2 so the job #4 will only have small vibs.

The Backup Copy Job is in 3 different locations because it's for Quarterly, Monthly, and Weekly archive.
That's how I get away from doing it with tapes because they are off-site.
Is that not the way to do it?

Thank you so much for your prompt responds all this time.

[Shestakov]

Glad we found the way to deal with the backup jobs.
For Quarterly, Monthly, and Weekly archive backup copy job has a GFS retention policy. You can use just a single job instead of 3 and save more repository space due to historical backup files deduplication(same restore point can be a monthly, weekly and quarterly). Please take a look.
Thanks!

[iamxCPx]

I see.
I think I know where you are going with that.

But the problem that I'm having is the limited space on the repository.
Each Repository is a 5TB USB drive attached to a PC at 3 different locations.
I think the ideal way to do the way you are suggesting is to have one giant 15 TB USB drive at 1 location.
I don't think I can plug in 3 x 5TB USB drive into one PC and call it one big giant repository, can I?

Thanks.

[Shestakov]

Since the .vbk size is about 300MB, your 5TB repository will keep up to 16 historical restore points. I`m not aware of your plans, but it might be enough for a while.
As to storage pool, I believe you can set something like Storage Spaces.
Thanks!

[iamxCPx]

The 300MB .vbk size is for the file server only. My whole environment .vbk is 1 TB.
So it can only handle 3 restore points. 4 is pushing it.

Storage Spaces is very interesting. Thank you for the link.
The problem is the remote PC OS is Win 7 Pro where the USB drive attached to.
Another thing that I'm worried about is data corruption when one drive goes bad. I might as well do a NAS at the offsite location.
Have you had any success with Storage Spaces?
I might have to do a little research.

[Shestakov]

I see your point. In a case of disaster with one of three USB Drives you may loose 1/3 of backups. Just like you can loose backups from NAS in the same case.
The best way to avoid it is to have several copies of the backups. There is also Veeam Surebackup option for the recoverability testing. Please take a look.

Have you had any success with Storage Spaces?

To say the truth, personally I do not have a lot of experience working with Storage Spaces, but it should work with no problems. Thanks!