• EN
Solved

Reorganization of Backup Plan to achieve close to 3-2-1-1-0 rule with our Infrastructure

  • 14 August 2023
  • 7 comments
  • 122 views
NemanjaJanicic
Userlevel 7
Badge +1

Hello to all Backup solutions architects 😀,

 

I will need advice in configuration of our Backups so that I can achieve as close as possible to the 3-2-1-1-0 rule.

Our Current Backup Infrastructure is as follow:
1 x Synology RS3621xs+ in Vienna, Austria.
1 x Synology RS3621xs+ in Novi Sad, Serbia.

1 x Synology DiskStation DS1019+ in Novi Sad, Serbia.

I wanted to implement Wasabi to our Infrastructure but didn’t get the green light, it is not approved.

 

Current Backups are setup in the following:
1. All our critical infrastructure is being backed up to the NAS in Vienna with this configuration:
I had an topic where we discussed about not needing Synthetic Full Backups periodically and Active Full Backups both enabled for the job. Need more opinions on this since I’m not professional in Veeam.

Current Configuration is set up like this:
 

 

  1. I created a Backup Copy Jobs on “Immediate Copy (mirroring) using the two WAN Accelerators. 1 is for our largest VM and 2 is used for other smaller size VM’s all together.
    Wan Accelerators are set up in High bandwidth mode.
     
  2.  For the third smallest NAS in Novi Sad I was doing manual copying of 1 Full Backup per month of each VM.
     

I was doing research and saw that 3-2-1-1-0 rule is the way to go.
What could I do with our current Infrastructure to optimize our backups as best as possible?
Is there an option to create Imutable backups to our smallest NAS with only Veeam software or with some Linux based tools since Synology is based on the Linux.
I would also like to implement SureBackup to achive the 0 errors.

Any advice would be greatly appreciated.
I know that its a total reconstruction and it has a lot of work to be done but I’m ready for it. 
Our throughput on each location is Gigabit connection.
Upload and download speed are 750 Mbps.



 

icon

Best answer by coolsport00 13 February 2024, 15:58

View original

7 comments

drews
Userlevel 3
Badge

Namanja,

Drew with Wasabi here - curious about your statement
“I wanted to implement Wasabi to our Infrastructure but didn’t get the green light, it is not approved.”

Is there something we can help with to get approval? Questions about security? Durability?

Please let me know how to help you.

Best regards,

Drew



 

NemanjaJanicic
Userlevel 7
Badge +1

Namanja,

Drew with Wasabi here - curious about your statement
“I wanted to implement Wasabi to our Infrastructure but didn’t get the green light, it is not approved.”

Is there something we can help with to get approval? Questions about security? Durability?

Please let me know how to help you.

Best regards,

Drew



 

Hello @drews,

 

CEO of the company didnt approve paying for Wasabi. I need to stick with current infrastructure that we have. That is the final decision. 😅

coolsport00
Userlevel 7
Badge +17

Hi @NemanjaJanicic -

Not sure I can answer all what you’re asking, but can assist a little bit.

I don’t know why you have both Synthetic and Active Fulls configured for any jobs. They both do the same thing (create a Full), albeit in a different way (via files on Repo vs pulling data across from source). Synthetic generally is the way to go here, especially if you using ReFS (on Windows) or XFS (on Linux) for your Repo as the process is almost neglible time-wise.

Immutability is definitely a recommended implementation if you can. There are many resources to help you out configuring this in your environment.

Lastly, for implementing the 3-2-1-1-0 rule...well, I think you have a good handle on this at this point. You are best suited to determine how best to have 3 copies of your data (1 being prod), on 2 media types (based off what your infrastructure/operational budget allows and what you currently have), with 1 copy offiste (you have backup copy jobs), and 1 either being immutable (I shared about this above) or air-gapped (i.e. tape?), and then 0 recovery errors by testing via Surebackup, which I believe you have a couple posts on already?

If you have further questions, don’t hesitate to ask..but I think this covers it 😊 

Shane Williford - Veeam VMCA/VMCE | Veeam Legend | VUG Leader | VCP-DCV | Twitter: @coolsport00
Geoff Burke
Userlevel 7
Badge +22

 I am assuming in the Synology you would create Linux VM’s with XFS for fast cloning etc. You could also then leverage immutability and create a hardened repository, but keep in mind virtual servers are vulnerable at the Hyper-visor level so much better to have physical servers. I know myself that trying to justify costs to Management can be difficult but you can remind them that the risks, especially with ransomware are very real and can be devastating to your business. When I worked for a service provider I heard a number of times “we should have just spent more on the right solutions and avoided this”. 

VMCA2022, VMCE2023, CKA, CKAD, KCNA, LFCS, PCA,TARS
coolsport00
Userlevel 7
Badge +17

Hannes from Veeam does a GREAT job on how to create an immutable server:

https://www.veeam.com/blog/installing-ubuntu-linux-veeam-hardened-repository.html

And, how to select proper hardware (if physical):

https://www.veeam.com/blog/hardened-linux-repository-best-practices.html

Shane Williford - Veeam VMCA/VMCE | Veeam Legend | VUG Leader | VCP-DCV | Twitter: @coolsport00
HunterLAFR
Userlevel 7
Badge +8

I’d say, with the resources you have,
If you want immutability, you can deploy a Linux Virtual Server, and mount in there the Synology Volume to be used for the Immutable backups, so in case of ransomware, your vm can be hit (with the hypervisor) but your immutable data will stay immutable and “untouched”.

Wasabi for external backups would be the way to go, but if you have a NO from CEO, nothing to say.

SureBackup, if you have limited resources, I did the following some time ago:

I setted up, and day by day, just with a couple of vms, taking into count dependencies, to be able to test my backups, is not a perfect solution, but worked for me and I was “sure” that I was able to recover.

If you have your plan draw somewhere, share it with us to take a look and give you some advice if needed.

cheers.

Luis F.- HunterLF - HunterLAFR - lfconsulting.org
coolsport00
Userlevel 7
Badge +17

Hi @NemanjaJanicic -

I just wanted to follow up on your post here again and see if there was anything further you needed to ask about your 3-2-1-1-0 post? If you feel one of the provided comments has satisfied your questions, we ask you please mark one as ‘Best Answer’ so others who come across your post with the same question may benefit.

Do not hesitate to ask further questions if needed though!

Thank you.

Shane Williford - Veeam VMCA/VMCE | Veeam Legend | VUG Leader | VCP-DCV | Twitter: @coolsport00

Comment


Terms & Conditions