Aggregate Reconstruction

[CLOUDMINER]

What is the best way to determine when aggregates should be reconstructed to improve performance?  What is involved in doing this?

[ekashpureff]

CloudMiner -

Another question without a short answer.

You're probably referring to recreation rather than reconstruction.

(reconstruction would be on a corrupted aggregate and is painful.)

When you know the aggregate geometry is out of guidelines, or when the disk utilization exceeds 50%.

Rather than recreate the aggregate you could add disk to it and optionaly run the 'reallocate' command on volumes in it and the aggregate.

That's the short answer...

🙂

I hope this response has been helpful to you.

At your service,

Eugene E. Kashpureff, Sr.
Independent NetApp Consultant http://www.linkedin.com/in/eugenekashpureff
Senior NetApp Instructor, FastLane US http://www.fastlaneus.com/
(P.S. I appreciate 'kudos' on any helpful posts.)

[CLOUDMINER]

OK great - thanks.  So when the disk utilization exceeds 50%, I would add disks and run the reallocate command, which would add the new disks to the aggregate.  Is that all I'm doing with reallocate, is just finalizing the process of adding new disks to the aggregate? 

When you say 50% - where are you viewing that and is that 50% of iops capacity?

[ekashpureff]

Cloud Miner -

Adding the disks to the aggregate adds the disks.

The 'reallocate' command is like defrag.

It will redistribute the existing volumes accross the new disks.

I hope this response has been helpful to you.

At your service,

Eugene E. Kashpureff, Sr.
Independent NetApp Consultant http://www.linkedin.com/in/eugenekashpureff
Senior NetApp Instructor, FastLane US http://www.fastlaneus.com/
(P.S. I appreciate 'kudos' on any helpful posts.)