• Status: Solved
  • Priority: Medium
  • Security: Public
  • Views: 1734
  • Last Modified:

netapp LUNs and Volumes

I have two netapp devices for which are configured synchronous  replication( snapmirror) on all volumes.
I am planning to change the replication for some of the volumes to asynchronous. Every Volume has one or two LUNs.
My question is how I should design LUNs space inside Volumes, and volume space inside Aggregate. How much space need to be reserved. For example my LUN is 500GB and my volume is 600GB. And if the changed data inside the lun is more than 100GB what will happen?
What is the best practices design?
0
dedri
Asked:
dedri
  • 3
  • 2
3 Solutions
 
dedriAuthor Commented:
any storage administrator here?
0
 
Paul SolovyovskyCommented:
Normally you want to have one LUN per volume, otherwise when you snapshot you snapshot both LUNs which may cause issues, especially when your LUN/volume grows due to snapshots, fragmentation, etc.. you cannot control each LUN individually as it is in the same volume with another LUN.

If you have a 600GB LUN and there are changes and you take a snapshot then not much happens since the snapshot is a point in time.  If the data is deleted or changed again then the snapshot will use some space.  On the other hand if a lot of data changes on a daily basis you may need to reclaim the space as it may get fragmented, this is where snapdrive comes into play.

Create a volume with a single LUN per volume.  Do use snap reserve (set it to zero).  You can reserve the space for the LUN or not (this will thin provision the LUN).  For a 600GB LUN I would create a 700GB volume to start (this can grow and shrink on the fly just keep in mind that if you grow it don't forget to grow the destination for snapmirror first).  If you thin provision it you'll see how fast it will grow over time and can adjust accordingly
0
 
robocatCommented:
Read the "fractional reserve" section of this document:

http://media.netapp.com/documents/tr-3483.pdf

By default a fractional reserve of 100% will be provided, needing 100% of extra space. This will probably waste a lot of space, so setting the fractional reserve to 0% and using the snapshot autodelete functionality may provide a better answer.

How much extra space should you provide ? It depends on the change rate of your LUN and the number of snapshots you keep. Mind you, if space runs out, the LUN will probably go offline, and you don't want that to happen. This could happen if the snapshots can not be autodeleted (eg. when using snapmirror/snapvault).

So start off by giving the volume plenty of space and lower this as you get a feel for the LUN change rate. As you probably want to keep some snapshots, make sure the snapshot autodelete is not triggered in normal circumstances.
0
Independent Software Vendors: We Want Your Opinion

We value your feedback.

Take our survey and automatically be enter to win anyone of the following:
Yeti Cooler, Amazon eGift Card, and Movie eGift Card!

 
dedriAuthor Commented:
if I configure fractional reserve to 0% and snapshot autodelete, do I have any grow of Volume. Do I have to make space reservation for aggregate, where volume resides and how much should be this aggregate.
Just to ask you regarding snaphost autodelete: My intention is to configure snapmirror asynchronous mirror for one of my LUNs on 2hr and on another LUN on 8hr. How many snapshot it will create - two and after creation of the second one it will  delete the first one( so I have to predict space for changes for two snapshots)?
0
 
dedriAuthor Commented:
I forgot to mention that my volumes are not flexvolume, If that matters.  I don't have such license.
0
 
robocatCommented:
(Flexvolumes are a basic functionality of any filer, but that doesn't matter)

To keep things simple, you should probably use one volume for each LUN, because you have different snapmirror schedules.

If you use async snapmirror, the latest snapmirror snapshot is always locked and can't be autodeleted. So any changes since the latest snapmirror transfer will require extra disk space in the volume. Make sure your volume is large enough to accomodate this for the worst possible case.

Unless you use thin provisioning of the volume, you need not worry about the aggregate.

Finally, there's also a volume autogrow feature. It can be set to make the volume grow by a certain amount when it runs out of space.
0

Featured Post

Granular recovery for Microsoft Exchange

With Veeam Explorer for Microsoft Exchange you can choose the Exchange Servers and restore points you’re interested in, and Veeam Explorer will present the contents of those mailbox stores for browsing, searching and exporting.

  • 3
  • 2
Tackle projects and never again get stuck behind a technical roadblock.
Join Now