cancel
Showing results for 
Search instead for 
Did you mean: 

Symantec Backup Appliance 5230

Hi,

I'm looking for some guidance/ technical assistance in implementing Symantec Backup Appliance 5230 within our backup environment.

 

Current configuration

Master Server: VM (RHEL 6)

Media Servers: Total of 4 physical media servers (2 in production and 2 in DR site)

Total Front End Terabyte (FETB): 80 Terabytes

 

Question #1:

If we replace all the physical media servers and replace it with Symantec Backup Appliance 5230, can we make do with two Backup Applince (Having one appliance in Production and the other in DR site)? Is this sufficient for us to be covered in a fail-over event (e.g. If the production backup appliance goes offline, how can the DR backup appliance be promoted as the Master server and how quickly can the fail-over happen)?

Also, if the production backup appliance goes offline and assuming that the DR site backup appliance resumes the role as the Master/Media server, can it still read data off the disk residing within production environment (assuming that the network connectivity is still available in prod)? Or is the storage on the backup appliance totally dependent on the appliance header to be online in order to function?

Question #2:

How much initial storage should we purchase/ allocate? Assuming that we need to protect an initial total of 80 TB of data. Is there a simple method on calculating/forecasting an approximation on the total storage require over a period of x number of years?

 

Any information/ guidance is much appreciated.

 

Thank you.

 

7 Replies

Question 1a: This depends a

Question 1a:

This depends a lot on the kind of jobs you're running, what the specs of the existing media servers are, etc.

As a general practice when I'm helping a customer replace existing media servers, I tend to do a 1:1 swap, unless the existing servers are just so old there is simply no comparison to the appliances.  Even then, I lean more away from consolidation than toward it.  Changing too many variables at one time is a recipe for failure in my years of experience.  Could you replace all of them with one appliance?  Maybe.  Should you?  That is a real 'it depends' question.  I'd suggest engaging your NetBackup partner SE for some sizing assistance to be sure.

Question 1b:

In a nutshell, you need the head unit to read the storage in today's world.  

I'd strongly suggest you look at AIR for what I believe you're trying to do, rather than having to promote media servers to masters and all the possible issues you could have with that.  With AIR, you have a master in each location and replicate the deduplicated data to DR via a Storage Lifecycle Policy (SLP), and immediately import the image into the catalog at DR, so none of those moving parts of catalog manipulation and server promotion are required.  The 5230 is capable of performing both roles (master and media) simultaneously, and in a small DR center, it could work to consolidate both roles -- but again I stress getting assistance with sizing.  AIR is the way to go here, though in my experience it drastically reduces the complexity of recovery, when time and simplicity matter the most.

Question 2:

The Symantec partner community and Symantec SEs have access to deduplication calculators that can help you with sizing that storage requirement.  The two most important factors that will influence this answer are the type of data you are backing up and what your retention periods for the data are.  Obviously the longer you intend to keep the data the more disk you'll need.  If this is primarily file system data that deduplicates well, and you have short retentions, you could see 99% deduplication on it and require very little physical capacity to meet the need.  If you have infinite retentions, even at 99% dedupe, you'll need more disk.

Hope that helps!

 

Hi John,   Little confused by

Hi John,

 

Little confused by your questions about the failover. You said you have a Master in VM, and you'll replace the 4 physical media servers with 2 appliances. So then there is no failover of the master, it still runs in the VM.

The amount of data to protect / the backup window = terabyte per hour required. This along with your retention requirement and the need for redundancy will determine the size and amount of appliance you'll require.

Your SE/Partner would need information about your backups set sizes for MS Exchange, File systems, HyperV, SharePoint (poor dedupe), MS-SQL, NDMP, Oracle, VMware etc.

You also need to determine how many instances you'd like to keep on the appliance e.g. Daily backups kept for 1 Month, Monthly backups kept for 1 year (+-42 instances).

Hi chashock, Thank you for

Hi chashock,

Thank you for your feed back. The backup type we currently have is a mixture of the following:

- Standard - Oracle RMAN, Solaris, HP-UX, Linux / Unix backup's.

- MS-Windows - Windows Servers + VM's backed up on folder level only.

- Flash-Windows / VMware - Standard VM backup's

- NDMP - CIF's shares

Data retention is configured as per below:

Daily Incremental = 4 weeks

Weekly Full = 6 weeks

Monthly Full = 12 months

Yearly Full = 7 years

 

Riaan, sorry I didn't mention that we also have a stand-by Master server which is also a VM. The consesus thus far seems to suggest that without engaging a Symantec Partner, working out the number of appliances required within the backup environment would prove difficult.

Assuming we keep the same data retention as we currently have on the backup appliance with a total FETB of 80TB, what would the storage requirement look like?

 

Thank you all for your assistance.

 

  I would go for what

 

I would go for what chashock suggested - with 4 NBU appliances, you can setup:

1 x Prod master server (appliance master mode, standalone)
1 x Prod media server (appliance media mode + additional storage shelf)

1 x DR master server (appliance master mode, standalone)
1 x DR media server (appliance media mode + additional storage shelf)

Added with AIR configuration (bi-directional), what is backup on Prod is replicated to DR and vice versa. 

In the case of 1 site goes offline, the restore capability for your previous backup is still intact (because of the replication), that's the best point of AIR.

However, if your concern is backup of clients in one site, then you need to setup cluster for your master server, this will be an added complexity, in this case you might have to get another standalone appliance for each site (for failover), or dont use appliance for the master server but with your own set of servers, just invest it onto the storage side (NBU appliance media server). My personal view.. there could be better options around from others.

For question #2, i am still trying to get myself familiar with the stacking options of NBU appliance storage shelf, so can't comment on that. It's best you get the SE to work that out together, I am sure they have some formula in place for your retention requirement.

It's not a matter of being

It's not a matter of being difficult to do, it's a matter of making sure you get it right the first time.  Moving to deduplication of any sort can create headaches you might not foresee if you aren't working with someone who's done it before, whether that's a SYMC SE or a SYMC Partner.

Not being familiar with the all the details of your situation, I'm not comfortable recommending any sizing to you over an internet forum.  You might be able to get that 80TB of data into a single 144TB media server appliance, but with those retentions and the disparity of data types you have, probably not.  There's design work to do so that you are successful in your deployment.

If you want to maximize your storage utilization, I'd recommend also making other changes to how you are backing things up.  I'd get rid of backing up VMs at a folder level -- use VADP backups for all virtual machines as it's typically faster, you get the file level granularity, and you can put limits on the resource utilization.  

I'd also consider at least testing the use of file system backups with Accelerator rather than using NDMP.  It'll provid you better performance, better deduplication rates, and more flexible restore capability.

Kindly advise us , our

Kindly advise us , our customer  will implement  Hyper-v High Availability so he will need Storage for this ,in the other hand he will use Backup Exec as Backup Solution .

 

We asked if he  can replace the original Storage (HP  storage ) with Symantec Appliance or not , and this will be good solution or not.

 

 

Please create your own post

Please create your own post if you want assistance. Also, this is not the Backup Exec Appliance forum. Please look here https://www-secure.symantec.com/connect/backup-and-recovery/forums/backup-exec-3600-appliance