Tek-Tips is the largest IT community on the Internet today!

Members share and learn making Tek-Tips Forums the best source of peer-reviewed technical information on the Internet!

  • Congratulations strongm on being selected by the Tek-Tips community for having the most helpful posts in the forums last week. Way to Go!

Poor Cloning Performance

Status
Not open for further replies.

morgaann

Technical User
Aug 16, 2002
47
CA
Hi there,

We have Networker 6.1.2 running on Solaris 8 backing up NT Clients. There are a couple of NT clients that backup about 150 GB combined data consisting of about 20,000 files combined each night and we not matter what we'vd tried can't get a clone to finish. The backup runs fine but the average throughput for the clone is 500kb/s.

We currently have these two clients in their own Networker group with their own Schedule and still the problem exists. We even tried deferring the clone using nsrclone and that didn't help.

Has anyone out there experienced this before and if so is there a solution.

Any help would be appreciated.

Thanks.
 
How many tape drives, and what type do you have?
 
We have an L700 jukebox with 7 DLT tape drives.
Thanks.
 
What is your server parallelism set to?

Remember, to do cloning, NetWorker has to read the data before it can then write the data.
 
my server parallelism is set to 42.

Thanks.
 

Are you running scheduled cloning or are you scripting the cloning?
 
I have actually tried both ways to clone - scheduled and deferred cloning and both ways result in the same problem.

Thanks.
 
I suggest that you try to determine which part of the cloning operation is having the performance problem.

Since cloning is essentially reading data and writing data,
Try to determine:

- how long does it take to save the backups
- how long does it take to recover the same saveset back to disk?

If there are no other backups running at the time you backed up thie saveset and then cloned, then:

1) total time for backup and cloning =
2) total time for backup
3) + time to read the backup and copy to clone tape

(approximately anyway).

You know what 1 and 2 is. What is the value of 3?

Once you determine these numbers, then compare and determine whether they correlate, or are the values too large, given the server parallelism, speed of tape devices, etc.
 
HI Morgaann

view yor Target Sessions per Tapedevice and set it to 1 because:
when you save more than one session per Tape in the same time Networker writes it im different Streams on the Tape maybe as shown here

( numbers represents Sessions )

1----2----1----3----2---3

when you Start cloning The networker writes The same Savesets on The Clonetape as Shown here

1-----1----2----2----3---3

so the Networker movs forward and Backward on teh Tape and you loose mutch time.

when you want to minimate these moves you must set The target sessions in every Device to 1

I Hope this will help you

cloner
 
Status
Not open for further replies.

Part and Inventory Search

Sponsor

Back
Top