Video Screencast Help
Symantec to Separate Into Two Focused, Industry-Leading Technology Companies. Learn more.

Netbackup AIR Configuration

Created: 02 Aug 2014 • Updated: 11 Sep 2014 | 9 comments
Ajith Sankar's picture
This issue has been solved. See solution.

Hi Folks,

  I implemented the AIR between my production and DR site. What i have done is listed below,

Production master server (Source) : Netbackup 7.6.0.2, Windows 2008

Production media server(Source) : Appliance 5230

DR master server(Target) : Netbackup 7.6.0.1, windows 2008

DR media server(Target) : Appliance 5220

Production and DR site connectivity band width is 100Mb/S (Not dedicated for backups)

Actually the DR Appliance is physically located in my production site and Replication is working smooth.

All the backup images are replicated from Production Appliance 5230 to DR Appliance 5220/

Now we are planning to move the DR Appliance 5220 to DR site physically. And the average incremental data is 1TB on daily basis.

And now my concern is how do i calculate the replication window, how much band width it will use ect things once i move Applinace 5220 physically to the DR site.

 

 

Operating Systems:

Comments 9 CommentsJump to latest comment

Riaan.Badenhorst's picture

Hi Ajith,

 

What you could do is look at something like this http://www.symantec.com/docs/TECH147351

 

Also discussed in this note https://www-secure.symantec.com/connect/forums/act...

 

What you could maybe also look at are the figures in this report from the appliance

http://www.symantec.com/docs/HOWTO94032

The space used for protection would be what changes each day as you replicate (although there would be some expiration as well)

Regards,

Riaan Badenhorst

You need an OpenVision to see the truth about Backups. Restores are a plus. But that's just Semantics ;)

ITs easy :)

Marianne's picture

Use this URL:

http://www.calctool.org/CALC/prof/computing/transfer_time 

Type in amount of data to be replicated on a daily basis and current link speed between sites.

For example - 1TB at a sustained rate of 100 Mb/sec will take more than 21 hours to replicate.

If you need to complete replication in less than 12 hours, you will need sustained throughput of close to 200 Mb/sec.

Supporting Storage Foundation and VCS on Unix and Windows as well as NetBackup on Unix and Windows
Handy NBU Links

Ajith Sankar's picture

What is the best way to find the total data size that replicate to the target domain after the deduplication.

In activity monitor i could see the Replication jobs and the data size(Kb) too. When i checked that for a day, it is near about 1 TB. Is that the actual data that is transferring from source to target(post dedupe) ? or is it a  prededupe data ? 

When i generated an Opscenter report for a day-->Dedeuplication report-->Deduplication report by master server, i could see below information

Master Server           Pre Dedup Size(GB)   Post Dedup Size(GB)   Deduplication Rate
  
Master.com               3,692.33                     191.64                          94.81  

This means only 191.64GB is the data that saving on the disk(source). In that case only this much data should tranfer to target.

Could anyone clarift this ?

 

Riaan.Badenhorst's picture

Hi Ajith,

Assuming the you transfer all of the backup for this master (or client) then 191.64 would be what is transferred. It is the unique data.

 

Regards,

Riaan Badenhorst

You need an OpenVision to see the truth about Backups. Restores are a plus. But that's just Semantics ;)

ITs easy :)

Ajith Sankar's picture

I observed that the data size which is displaying in the Replication jobs are just the pre-dedup data. i could see the dedup average rate is 85%, and total data size of all replication jobs are 1000GB.So in this case, i can assume that the150GB of data would be transfering from source to target.

Hope this is correct.

Riaan.Badenhorst's picture

Hi Ajith,

 

Sorry but the figures you are quoting keep changing. Please remember we can't see your, or have knowledge of your, environment.

 

In a job it would say "CR sent", and that is what is actually transferred. The value is in the Kbytes tabs are what is protected i.e. prededup.

Regards,

Riaan Badenhorst

You need an OpenVision to see the truth about Backups. Restores are a plus. But that's just Semantics ;)

ITs easy :)

SOLUTION
Ajith Sankar's picture

Hi 

I have calculated the CR sent data for few replication jobs. And i found it is giving the same figures what i calculated earlier.

Marianne's picture

It seems you have your answer.

'CR sent data' is the actual amount of data to be transferred.
Monitor the environment for at least a week to get a good idea of what happens for full and incremental backups.

So, this means that bandwidth requirement is a lot less than you initially thought.
You can still use the calculator in my post above.
 

Supporting Storage Foundation and VCS on Unix and Windows as well as NetBackup on Unix and Windows
Handy NBU Links

Riaan.Badenhorst's picture

Great, problem solved wink

Regards,

Riaan Badenhorst

You need an OpenVision to see the truth about Backups. Restores are a plus. But that's just Semantics ;)

ITs easy :)