One of the coolest features of ExaGrid is its capability of setting up specific algorithms on a given share we create on it, based on the backup software it is going to work with.
In this way, a user can be sure he can take full advantage of every single bit of space on the ExaGrid for his backup solution. The list of supported software comprehends basically the most common backup solutions, as you can see here:
In this article, we are going to test the use of ExaGrid in conjunction with Veeam Backup & Replication.
Creating a share in the ExaGrid is a short and easy task, independently of the backup software we choose.
We go in the Manage -> Shares section of the web interface, where we can see a list of already configured shares.
You can create different shares for different backup types, all on the same unit.
We then open the Share Creation interface by hitting the “New…” button.
ExaGrid needs to know the Share Type, if it is going to be published via CIFS or NFS protocol (OST in not an option for Veeam shares), and the name.
Also, we need to set the IP address of the Backup server that is going to use this share. In my lab, 192.168.0.221 is the Veeam Backup Server.
Once created, we can see the share in the share list:
With these few settings, ExaGrid was ready to accept Veeam backups.
ExaGrid and Veeam
Veeam is, in my opinion, the best vmware backup software available today. Fast, reliable, full of incredible features and ease of use at the same time. ExaGrid has specific settings for Veeam bakups, and they also give you recommended settings to use Veeam at its best.
So I was incredibly curious to test Veeam in conjunction with ExaGrid.
My Lab is a replica of a complete production environment: there are DCs, a file/print server, an Exchange 2003 server, an ERP, and some other VMs.
This is a report of the existing servers we are going to backup:
As you can see, there is a fair amount of data, mixed guest OS, and different vmware virtual hardware version (so different behaviour regarding CBT).
My Veeam Backup Server is a Windows 2003 R2, 64 bit, installed as a virtual machine (2 vCPU, 4 Gb of vRAM) inside the same cluster we are going to backup.
I created a single backup job for every server I’m going to backup: in this way when you need to change schedule, retention, or other parameters, one modification does not affect other VM’s backups.
ExaGrid has done a great work in creating guidelines for configuring every supported backup software, and you can download directly those pdf files from the management interface.
I read carefully the Veeam Guide, and the backup configuration I’m goind to show you is to me the best fit for Veeam + ExaGrid.
First of all, every Windows server is going to write to ExaGrid, needs to be modified with a registry key, needed to increase the CIFS writes timeout key. The key too is available on the Online Library into the ExaGrid, as the Veeam Guide.
On the Veeam side, every backup can be configured as desired, but keep in mind these configurations for best performance:
- Traffic encryption affects backup speed, but not deduplication speed. Use it if needed.
- AVOID Reverse Incremental. With this type of backup Veeam is going to do post-processing on the backup set to create the reverse incremental backup set (please refer to the Veeam Backup User Guide to understand what it is). This activity conflicts with the ExaGrid post-backup processing, reducing the effectiveness of ExaGrid.
Use Incremental instead as displayed:
- On the storage tab, use these settings:
Setting compression level to none and optimization as “local target” are voted to one goal: let Veeam go at full speed without loosing time compressing data. In this way, we can finish backup activities as soon as possible, and then use ExaGrid deduplication to re-gain space.
- Ensure that “Enable automatic backup integrity checks” is checked.
My first backup towards the ExaGrid was finally launched:
We will see in the next articles the results.