Hi,
Excellent news about v6 becoming GA hopefully later this week. With this is mind, I'm preparing to perform a redesign of my backup environment to make better use of the new functionality.
Currently, under v5, we have two physical Veeam backup servers, backing up approximately 125 virtual machines each to their own SAN storage presented to the servers using FC. This means that Server A has its backup repository on Storage A and Server B has its backup repository on Storage B. Both servers are using their own database on a shared SQL instance.
Under v6, we are looking to redesign the environment. I am thinking of creating a virtual Veeam Backup Manager server to manage the backups. The two existing physical servers would then become proxies for the Backup Manager. My concern is about the SAN storage presented to the physical servers.
Should the storage still only be presented to the one backup proxy, like in v5?
If this is the case, does the Backup Manager know on which server it has stored the backups to allow quick recovery?
If a proxy server fails and the Backup Manager chooses another proxy to perform the backup, does this mean that it will perform a full backup or still continue with incrementals, reversed incrementals etc.?
If you have any documentation (e.g. the User Guide for v6) available, then I'm sure that these questions will all be answered in there, and it would be handy to have this available early to allow us to formulate the design and implement it once the new release is available.
Thanks in advance for your assistance with this.
Chris
-
- Enthusiast
- Posts: 29
- Liked: 1 time
- Joined: Nov 26, 2009 11:46 am
- Full Name: Chris Bell
- Contact:
-
- Chief Product Officer
- Posts: 31772
- Liked: 7274 times
- Joined: Jan 01, 2006 1:01 am
- Location: Baar, Switzerland
- Contact:
Re: v6 Backup Proxies and Repositories
Hi Chris,
Did some cleanup to delete previous incorrect/confusing response by Alexander.
Essentially, you do not present your storage to the backup server - you now present the storage to backup repository, as this is the component that does all the backup target storage interaction. Backup proxies and backup repositories are two separate components in v6, they are completely independent from each other (although both can potentially reside on the same server). You do not need to "present" backup repositories to backup proxies either - any backup proxy can write any backup repository over the network, so don't worry about that. If it happens that both proxy and repository are sharing the same server, then of course network connection is not needed.
Because backup repositories are accessible over the network, any proxy can get access to the backup files stored in the repository, despite the fact that actual repository can be backed by local or direct attach storage on that server (meaning that directly, the storage is only accessible by this specific server).
Each time the job runs, every VM can potentially be assigned to a different proxy server than it was processed with on the previous job run (depending on environmental conditions). This, of course, does not affect the job itself - it still continues with incrementals.
Let me know if the above is clear.
In addition to regular documentation, I am working on extensive FAQ to make it available along with the release, this should answer a lot of these questions.
Thanks!
Did some cleanup to delete previous incorrect/confusing response by Alexander.
Essentially, you do not present your storage to the backup server - you now present the storage to backup repository, as this is the component that does all the backup target storage interaction. Backup proxies and backup repositories are two separate components in v6, they are completely independent from each other (although both can potentially reside on the same server). You do not need to "present" backup repositories to backup proxies either - any backup proxy can write any backup repository over the network, so don't worry about that. If it happens that both proxy and repository are sharing the same server, then of course network connection is not needed.
Because backup repositories are accessible over the network, any proxy can get access to the backup files stored in the repository, despite the fact that actual repository can be backed by local or direct attach storage on that server (meaning that directly, the storage is only accessible by this specific server).
Each time the job runs, every VM can potentially be assigned to a different proxy server than it was processed with on the previous job run (depending on environmental conditions). This, of course, does not affect the job itself - it still continues with incrementals.
Let me know if the above is clear.
In addition to regular documentation, I am working on extensive FAQ to make it available along with the release, this should answer a lot of these questions.
Thanks!
Who is online
Users browsing this forum: Amazon [Bot], kurtis, Majestic-12 [Bot], Regnor and 86 guests