Backup Exchange located on Azure VM

[Jrouzies]

Hi everyone,

I need some advises about the best backup configuration I could get with the following architecture:

• 4 Exchange servers on Azure (VMs)

• 1 Veeam B&R server on Azure (VM)

• Backup storage is direct attached disks to Veeam server

The Exchange database distribution (simplified):

Servers : | 1 | 2 | 3 | 4 |
--------------------------------
----DB01 | A | P | - | - |
----DB02 | - | A | P | - |
----DB03 | - | - | A | P |
----DB04 | P | - | - | A |

A = active ; P = passive

(etc, there is 16 DBs in total).

• DBs are on same volume on each server.

• We have ~10 TB of non-replicated data

Now what are my problematics:

• I cannot backup the VM, because well Azure.

• I cannot backup DB volume, because I would be backing up Active and Passive at same time, resulting in loss of storage

• > I feel stuck with File-Level Backup, and excluding Active DBs path. Is it a bad things? It says its the slowest backup mode. Any better alternative?

We were using OnPremise VMware ESX servers, and backup was crazy good: ~ 10 TB backedup in few hours, compression or dedup (not sure) was 50%, so we could store 2 backups in roughly 11 TB.

Today with the file-level backup:

• compression is nowhere to be seen

• performances for full-active backup are okay (~60 MB/s)

• performances for incremental backup are a disaster(~10 MB/s) > so we ended up doing only full backups so far, meaning we go from 11 TB storage to ~25 TB storage because no compression and to have spare during backup rotation

If anyone knows a better alternative, I'm all ears, pretty new in the backup domain.

Thanks!

[Mildur]

Veeam Agent for WIndows has full support for Exchange DAG. This is Block Level (Volume Level) with Veeam AGent for WIndows Product.
You don't have to create file Level Backups anymore.

https://www.veeam.com/kb2463

Microsoft Exchange Database Availability Group
Starting with Veeam Backup & Replication 9.5 Update 4, Microsoft Exchange database availability group (DAG) nodes are now automatically processed in a sequential manner, so it is possible to use Veeam Agent for Microsoft Windows to process Microsoft Exchange Database Availability Groups (DAGs).

The procedure of adding a DAG to a Veeam Agent backup job differs depending on the type of the DAG that you want to process:

For a regular DAG, the backup job configuration procedure is the same as for any failover cluster mentioned above, so all the steps above are relevant
For an IP Less DAG (a DAG without an Administrative Access Point), the backup job configuration procedure is similar to the same procedure for standalone servers. To process an IP Less DAG, you must create a Veeam Agent backup job for servers in the Veeam backup console and add all nodes of the IP Less DAG (the steps below show it in detail) to this job.

[Jrouzies]

Hello Mildur, thanks for your reply.

Sadly I already searched and found the option of Windows Backup Agent, but from what I read, they backup BOTH Active and Passive databases, with no option to exclude:

veeam-agent-for-windows-f33/physical-ex ... 01-30.html

Hi,

Regarding backup option to only backup the passives for DAG deployments is that feature in v10?

Best regards!
Victor

Hello Victor,

Unfortunately this feature wont be included in the upcoming v10 but the suggested workaround remains valid. Thanks!

More recent validation:

Hello Samirb, welcome to the community!

You cannot back up only one DB copy with the current version of the Agent.
I will count your post as a vote for the feature to be implemented.

Thank you!

Which means double storage for nothing. Unless i'm missing something?

[Mildur]

Ok, that‘s bad.
You are right. You have to backup both databases. Passiv and active.
Because veeam agent Jobs has backup files per agent, there will be no compression (veeam „dedup“ of similar blocks in a single job)

Do you have the possibility to run one of the exchange server with only passiv copies of all databases? Then you can backup only this passive dag member with veeam Agent for windows.

[Jrouzies]

Yes I don't understand why they don't implement exclusion of Passives especially they seem to be able to detect them:

"(only passive database copies are processed, active ones are excluded from VSS freeze operations; however, the job still copies all the correspondent EDB-files of all databases from all the nodes)."

I'm aware of the full-passive node easier to backup, but it would only make us pay one VM that will be standby 95% of the time. And this is supposing we remove any Passive DB from the other servers, if we add this server on top of the rest, then we pay our storage of 10 TB of non-replicated data again.

It feels sad that we don't have better exclusion system. On the backup directly at the VM level, or even at storage level we used to have, we could exclude specific paths, and it was still really fast block-level backups.

[Jrouzies]

Is it true that Veeam can do some sort of Dedup on the fly during the backup?

Refering to those posts:

• veeam-backup-replication-f2/backup-larg ... 37048.html

• veeam-backup-replication-f2/read-this-f ... tml#p95276

Could it mean that Veeam would be able to dedup the data between Active and Passive if following with the Veeam Agent backup method?

Thanks

[foggy]

If we're talking about Veeam B&R (not Veeam Agent for Windows), then it will dedupe between VMs considering the repository is not set up to store per-VM backup chains.

[Mildur]

Is it true that Veeam can do some sort of Dedup on the fly during the backup?

As Foggy said, this is true with Veeam B&R Image Level Backups (vsphere/hyperv Integration).

You have to use Veeam Agent for Windows to backup your Data on Azure VMs. No „veeam dedup“ for that. ;(

[Jrouzies]

Thank for you replies.

What about Windows Dedup?

As far as I saw, it seems to dedup only after the backup is done and some scheduled tasks run. Or once it did, can it dedup on the fly?

[foggy]

Windows Data Deduplication uses a post-processing approach.

[Dima P.]

Veeam Agent for Windows to backup your Data on Azure VMs. No „veeam dedup“ for that

Deduplication works between the data within a single job, so when we talk about workstation or server agent backup deduplication will have minimum impact (since each agent machine has it's own backup job). For failover cluster jobs all nodes are placed within single backup file, so you may notice decent deduplication ratio.