Highlighted
boerbokrib
3 Argentium

Staging large volume

Jump to solution

Hi

I have a large volume of 38 terabytes that i wish to stage to another volume on the same data domain.

I have run a command line nsrstage with -f option reading a file of all the save sets . however i am only getting about 34 gb per hour throughput.

The command i have used is.

nsrstage -m - v -b clonepool -S -f and file with output of volume I want staged.

is there a way to check where the bottle neck is and is there another way to do this that is quicker without saveset id changing?

0 Kudos
1 Solution

Accepted Solutions
bingo.1
4 Beryllium

Re: Staging large volume

Jump to solution

Look at this example ...

stage_1.jpg

This has been done with a NW 8.2.0.0 server/SN on a DD880 running DDOS 5.5.0.4 (as mentioned).

Command was "nsrclone -m -b <backup_pool> -S ssid/cloneid"

As we do not have a remote SN in this environment, i could not make any mistake by specifying the wrong source and/or dest. storage node. Network traffic is 0.

View solution in original post

0 Kudos
9 Replies
bingo.1
4 Beryllium

Re: Staging large volume

Jump to solution

I assume that your sources reside on another clone media.

If so, do not forget to specify the ssid along with the cloneid (ssid/cloneid) - otherwise you will end up with only one instance.

For the speed, I am afraid that nobody can help you as we do not know your environment like device/media types and whether you stage locally (on one storage node) or remotely (if the data has to travel between different nodes).

 

However, you could split your save set file and run multiple save set commands at the same time. It would be great ifyou clone from multiple media at the same time.

Of course, the reason could also be a software bug.

ble1
6 Indium

Re: Staging large volume

Jump to solution

I did recently this as I was migrating data from old pools to new ones on same and remote DDs an I simply used:

nsrclone -m -b <target pool> -d <destination storage node> -S -f <input file>

For local one, I didn't pay attention to stats, but I'm quite sure what you had in 1h I did have in 1 minute (DD890).

boerbokrib
3 Argentium

Re: Staging large volume

Jump to solution

Hi it is a dd860 the volume fsoff.001 resides on this dd860 and i want to stage it to the volume fsclone alos on the dd860.

The reason for staging is i want to do away with the fsoff which is a pool that was used in the past and no longer is functioning.

I was hoping that as this was all on the same ddomain it would be fast we have sata drives in the archive where this volume resides due to it all being old data.

0 Kudos
crazyrov
4 Tellurium

Re: Staging large volume

Jump to solution

Is the devices configured as a DD device or as an AFTD ?

0 Kudos
boerbokrib
3 Argentium

Re: Staging large volume

Jump to solution

Hi They are configured as dd devices.

0 Kudos
boerbokrib
3 Argentium

Re: Staging large volume

Jump to solution

Hi I have also found that my archive tier is full on the dd860. WHat effect if any could this have?

0 Kudos
bingo.1
4 Beryllium

Re: Staging large volume

Jump to solution

No idea about the archive as we do not use it (yet).

However, i just migrated about 1.6 TB on our DD880 yesterday (DDOS 5.5.0.4) and it was finished within minutes.

boerbokrib
3 Argentium

Re: Staging large volume

Jump to solution

My networker server and storage node sit across a wireless link from my datadaomain 860.

Will this affect the nsrstage as it is copying the data back to the storage node before copying it back to the datadomain?

0 Kudos
bingo.1
4 Beryllium

Re: Staging large volume

Jump to solution

Look at this example ...

stage_1.jpg

This has been done with a NW 8.2.0.0 server/SN on a DD880 running DDOS 5.5.0.4 (as mentioned).

Command was "nsrclone -m -b <backup_pool> -S ssid/cloneid"

As we do not have a remote SN in this environment, i could not make any mistake by specifying the wrong source and/or dest. storage node. Network traffic is 0.

View solution in original post

0 Kudos