Cloning size Issue

I do not completely understand. It obviously looks like more data is transferred than the actual save set size. I never looked so carefully to these numbers but that would indicate to me that the tape drive itself has to retry a lot. This would also mean that the clone process would take much longer than usual. You would usually become aware of such issue. It could also result in a (much) smaller tape capacity.

If so, this indicates to a problem with your tape hardware. Please make sure that it runs fine.

If the duration is fine, my first guess would be a software bug. To avoid any 'trivial' NW issue, may I suggest that you upgrade to one of the current NW versions.

You will never lose the NW databases.

In general, NW upgrades work the same like with any other software:

  -  stopping NW and removing the current software (leaving the databases behind)

         You only need to delete the directory tree if you now want to completely remove the software

  -  installing the new version and restarting the daemons/services in the end.

May I suggest that you train the procedure as follows

  -  take a server and install your current NW version (it runs in eval mode for 90 days)

  -  run some backups using an AFTD disk

  -  verify the backups with mminfo & nsrls

  -  upgrade NW as described above

Whether you need a new license of course depends on your service level. Technically, there is no additional license required.

I think my software needed to  be upgrade to new versions, but our huge data size and risk of upgrade afraid me, my product 9.1.0.6 have licensed ...

do the upgrade process need an additional license ?

please guide me about correct upgrade software without losing data and database....

thanks alot

Thanks alot 

for example my backup size is 4tb , in clone process the size continuously increased from 4tb to 10tb and more and more to 42tb!!!! 

(42tb clone size instead 4tb backup)!!!!!!

Absolutely weird. And of course such dramatical increase would take much longer (and will use more tape space). However, if these values remain in a 'normal' range, then this is just bogus info.

Just remember that NW is not to blame - NW does not install its own device drivers but its own jukebox drivers.

The only detail we know is that some of your clones grew 10 times larger than the backup size. This will roughly result in a 10 times longer clone process with 10 times more tape consumption. However, this does not seem to worry you at all.

But what you see obviously does not appear on certain backups. So where/when does it occur:

  -  for a certain tape brand?

  -  for a certain tape drive?

  -  for a certain storage node?

  -  As you run NW 9.1.0.6 this installation have already run for while. If this is a new problem, what has recently changed?

Up to now we have just theroretically discussed how such behavior might occur. To give you more specific answers for a system we do not know, any further detail would be helpful. Time to improve ...

Hello

my question is why some of clones behave increasing size but others not!!!!

is this for configuration? 

or?

* my brand of Tape devices is Quantom scalar 6000

* all of tape drive not certain drive can occur this problem

* this problem is not new for our networker

* A ,B ,C ,D are physical servers

* our design is  A as root(mgmt-storage node-networker server)

 sub layer:                 B as a (networker server - storage node)

 sub layer:             

                                 C as a (networker server) + D as a (storage node) for C  [ C have huge traffic and files like database] and we separate C and D [before C and D was on same physical server]

this problem occur for storage node B and D  clones....

the size and time consumption create a problem with operator in my sub layer parts , that they want to troubleshoot this issue by administrator....

We are still not talking the same language.

'Scalar 6000' is the type of the jukebox, not the tape drive. Of course, a jukebox may behave wrong, but the 'loader' for sure has no influence on the tape characteristics. I guess, there are LTO drives installed. How many do you have?

I do not know what you mean by 'sub-layer' as this term ist totally unknown by NetWorker. It seems that you mean  NW 'data zones'. Obviously you run 3 NW servers (for A, B, and C) which share the jukebox. The tape drives as well?

C must also be a (local!) storage node because the bootstrap can only backed up to a local attached device. There is no way to avoid that.

So how are the tape drives connected/shared among the hosts? - a picture would h

BTW - which OS do you use for your hosts? - Did you ever try to test the tape drive with an OS-specific backup tool just to verify whether the issue occurs here as well?

Hello

- our scalar 6000 has about 2500 usable LTO7 tapes

- Site 1 we have 3 NW Servers that they backup to Data Domain 9500

- Site 2 we have 3 storage node for cloning on scalar 6000

- All of hosts installed windows server 2012

- 2 Networker Servers B & C occur this issue, some of clone size not all increased....

Does this picture come somehow close to your scenario?

Hello

thank for your design...

yes , our scenario is like this....

My pleasure. Such a simple diagram provided by yourself could have avoided lots of discurrsions.

So - if the problem 'only' occurs on 2 of your data zones, it would be worth to investigate, what is the difference between those and the working one. It there is none, I expect a hardware related problem.

Next I would verify whether the issue happens for specific tape drives (and tape media). If possible, swapping tape drives (even among data zones) could help to isolate the issue. However, I wave no idea how difficult that would be.