Start a Conversation

This post is more than 5 years old

Solved!

Go to Solution

2092

August 12th, 2010 22:00

Is it possible to change the NDMP Snapsure default storage pool for checkpoints?

The customer has a NX4 with 2 storage pools: one for SATA disks, and the other one for SAS disks.

There is a huge filesystem (2.7TB) over the SATA disks, using all the available space from that storage pool, so there is no space left. There is around 400GB of free space in the SAS disks.

The NDMP.snapsure parameter is set to 1. The backup server is running CA BrightStor ARC Serve (but the customer also tried with Veritas Backup Exec).

When we run an NDMP job to back up the filesystem over the SATA disks, it would fail saying that it can't create the snapsure filesystem. Taking into account this thread: https://community.emc.com/message/383367 , I manually created a checkpoint on the SAS disks storage pool. After that, we ran the NDMP job again, the automatic checkpoint was created on the SAS disks storage pool, and the backup went without problems.

It seems that when an NDMP job is run to back up a filesystem in a given storage pool, the NX4 tries to create the checkpoint in the same storage pool. If there isn't enough free space on that storage pool, the checkpoint creation fails, and the NDMP backup aborts.

The question is: is it possible to change the NDMP automatic checkpoint default storage pool? is there any way to tell NDMP to use another storage pool for the automatic checkpoint? Or maybe globally change the Snapsure storage pool behaviour, in order to try to pick up another storage pool if the default is full?

Thanks!!!

Gonzalo

August 12th, 2010 23:00

Hi Gonzalo,

It is not possible to change the NDMP default storage pool for check points.

The only solution is to go by https://community.emc.com/message/383367. That is to ensure the first check is created manually on the desired storage pool.Which will in turn use the selected storage for automatic NDMP check point.

There is no parameters to change the snapsure behaviour.

There is also no way to specify to use anaother storage pool when one pool is full.

1 Rookie

 • 

95 Posts

August 13th, 2010 02:00

Hi Gonzalo,

I have the same problem in another customer, and this is the response from EMC Support:

ME:

Ok, but customer doesn’t want to create manually a checkpoint for that FS. Customer created ckpt manually in the other pool and worked, but after delete the manual ckpt, the backups are failing again….

SUPPORT:

Once you create one checkpoint manually on the clarata_archive pool, from then on, all checkpoints should go to the clarata_archive pool.

So if you manually create a checkpoint for FS39 on the clarata_archive pool.  Then all of your automatic checkpoints should be created on that pool.

The other option to avoid creating them manually would be to create a schedule to create checkpoints.

As long as there is one checkpoint for that file system on the clarata_archive pool, all future ckpts will be created on that pool.  When you delete that ckpt the future ckpts will go back to the default pool.

This is the only way until now.... maybe we can do "Request for enhancement"...

Regards,

Dario

IPM

8.6K Posts

August 13th, 2010 02:00

just create a manual checkpoint or a schedule specifying where the savol should go - all other checkpoints for that fs will use the same savvol

2 Posts

August 13th, 2010 08:00

Thanks to everyone!!

I was almost sure that the only way was to manually create first other checkpoint on the desired storage pool, but anyway I wanted to give it a try...

I'll try to schedule a checkpoint right before the backup window.

Thanks again for the quick answers!!

No Events found!

Top