vault takes to long with one of two drives

[ruffy]

Hi together,

my problem is, i have nbu 6.5.1A with ADIC/Quantum Scalar i500 Tape Library and 2 drives. When my vault begins it creates two duplicate jobs. But after 5 hours is one of them finish and another one takes over 14 hours. Why nbu can not divide all of backup images to even number of images for both drives? Or what i have to do to configure it?

thanks for help

ruffy

[ruffy]

Has nobody any idea?

[Ravindra_2]

Hi

      Answer of this question is little bit tricky because you have not defind many things which can create impact on vault process.eg

 how many clients you have

what multiplex setting is done on policy  and specially on storage unit side

On what basis you are doing vault (Vault config)

whether you have backups active at the time of vaulting.

[ruffy]

Hi Ravindra,

thanks for your answer. Now i know which information do you need.

- i have 38 clients
- multiplex in the vault profil is not set (storage unit are the drives of the library)
- one policy for daily vault and weekly
  daily: Policy type is vault and in backup selections do i have "vltrun daily"
  vault profile: removable media and/or disk =2, no preserve multiplexing, duplicat smaller images first, and expre original disk backup images after 1 hour

do you need more informations?

[Ron_Cohn]

Ruffy,

The issue is not with the Vault process.  Is your initial backup to a VTL or to physical tape?  How many streams are going to each tape?  What is your multiplexing values.  I'll explain based upon your response.

[Ron_Cohn]

Ruffy,

In your response, you selected not to preserve multiplexing.  In this case, you will be making multiple passes thru your input.  Also, if you are using disk OR tape, it is best to set a fragment size so that NetBackup can quickly position itself without having to scan thru the entire input.

[ruffy]

Hi Ron Cohn,

thanks for your help!

My backup is to physical tape with two streams. One for each drive. I don't have multiplexing values, becaus i don't use multiplexing on vault. Is this function "preverse multiplexing"?

[ruffy]

Ron Cohn,

where can i configure a fragment size for vault job?

[Ron_Cohn]

Ruffy,

Can I assume your primary (or initial backup) is to disk?  Then you use NetBackup Vault to create a physical tape?  Anyway, under the NetBackup Administrator Console -> Storage Units is where you set the fragment size.  Thru the various manuals and discussion groups, we have decided to use 10240 (10 GB) as our fragment size.  In the mainframe world is was known as a tape mark...

[ruffy]

Ron Cohn,

at first thanks for your help. 

You are right, my primary backup is to disk. Then i use Vault and duplicate the B2D images to tape. I think it is not the problem of frament size of a storage unit, because the vault has nothing together with the originaly storage unit of a backuped client.

The problem is still actuall. I have two duplication child jobs. One of them has two images and the other one all another. Any ideas???

ruffy

[Ron_Cohn]

---

@ruffy wrote:

 I think it is not the problem of frament size of a storage unit, because the vault has nothing together with the originaly storage unit of a backuped client.

The problem is still actuall. I have two duplication child jobs. One of them has two images and the other one all another. Any ideas???

ruffy

---

Ruffy,

The Vault process has everything to do with how images are stored on the VTL - especially if multiplexing to the VTL.  For example on the VTL, tape #001 contains image_001, image_002, image_003.  Tape #002 contains image_001, image_003, image_004.

When you start the vault process, it goes to tape #001 and gets the image_001.  BUT, it knows that image_001 is also on tape #002.  It will dismount tape #001, mount tape #002 and continue vaulting image_001.  Once that is completed, it dismounts tape #002, mounts tape #001 and proceeds to image_002.  Since it is contained strictly on tape #001 - no problem.  Then it vaults image_003.  Again, it performs the same process as it did for image_001.  This is what will kill you when vaulting.  There is a fine line in trying to optimize writing to the VTL AND using Vault.

Because of the possible mounts / dismounts (even virtually), fragment size is very important.  If you did set fragment size, then going back to my example, when mounting tape #001 to begin vaulting image_002, NetBackup is going to have to scan tape #001 from the begining to get to image_002 starting point.

[Ron_Cohn]

---

Because of the possible mounts / dismounts (even virtually), fragment size is very important.  If you did set fragment size, then going back to my example, when mounting tape #001 to begin vaulting image_002, NetBackup is going to have to scan tape #001 from the begining to get to image_002 starting point.

---

The sentence should have read:

"If you did NOT set fragment size..."

[Omar_Villa]

Instead of vaulting why you dont try using Storage Lifecycle Policies which is the same thing runs duplicates and is even better to maintain and manage because you have all your retentions over a single policy this way you can see if is a drive/stu problem or if is the vault.

hope this helps.

regards

[ruffy]

@Ron Cohn:

thx for this description. But i did not set the fragment size till you say it to me. After i have set it. But with no cuccess!

And with standard configuration should your sample not to refer?

@Omar Villa:

i didn't know this funkction till you say it to me. I should read some docu and test it before i can use it. And i should have any solution for yet. Thanks anyway.

Message Edited by ruffy on 07-22-2008 11:38 AM

Message Edited by ruffy on 07-22-2008 11:39 AM

[Ron_Cohn]

---

@ruffy wrote:

@Ron Cohn:

And with standard configuration should your sample not to refer?

---

Ruffy,

I am trying to understand the above sentence.  Also, how many "duplication" steps are in the activity log when vaulting?

[ruffy]

Hi Ron Cohn

you've right. i have two duplicates (each for for one drive).

sorry for my english, when you dont understand what i mean.