NetApp - cloning volumes

Discussion in 'Networks' started by zebulebu, Mar 28, 2008.

  1. zebulebu

    zebulebu Terabyte Poster

    3,748
    330
    187
    Guys

    Bit of a long shot here I know, but has anyone here used or is familiar with NetApp SANs?

    I have one at my new job and have never supported them before. We had a potential issue arise yesterday where the US support guys weren't available and I may have had to do a refresh of a development environment on my todd. It was a reasonably hair-raising moment for me - I've not yet been on any training and, other than this, they pretty much manage themselevs so haven't needed to dive in and learn how OnTap or snapshotting works. Luckily, the need for the refresh went away, so it won't be an immediate problem for the time being, but I would like to get a headstart on a few things before the issue arises again.

    I understand how the technology behind them works, but the CLI reference for OnTap is a bloody nightmare to wade through, and seems to contradict itself at various points! I'm hoping that someone here will have supported them and be able to answer a couple of questions for me.

    Firstly, the environment is set up as follows:

    FAS270, single shelf of 13 144Gb disks with one hot spare, RAID6, single aggregate of 1.2Tb, OnTap v 7.1
    Eight volumes on the aggregate - various sizes, assigned to various DB functions (data, logs, indexes)
    Several LUNs, mapped via iSCSI to various disks on two servers (one dev, one production)
    iSCSI Vlan presented on our core switch, using QLogic HBAs
    Snapshots are created on a defined schedule (hourly for most LUNs) using SnapDrive on the two servers

    The problem I face is that I need to recreate the LUNs that are referenced by the dev server so that they now reflect current live data. As far as I understand the process, it is a matter of:

    Closing down the database and stopping anything that may be accessing the LUNs
    Deleting all snapshots of those LUNs that need to be refreshed
    Deleting the LUNs themselves
    Cloning the LUNs from the live data (either by cloning the LUN itself or taking a snapshot of the live data and Bringing that up as read-writeable)
    Bringing those LUNs up again by setting the iSCSI targets in our SCSI initiator software
    Bringing the database up again

    Now, as far as I understand it, there are different ways to go about cloning a volume - the two main ones are by 'split cloning' it - which creates a complete copy of the data (and therefore doubles the amount of data space required) or by cloning from a snapshot (which doesn't, as it just creates pointers to the data and therefore the only space necessary is that required to take snapshots of changed data)

    The CLI syntax is not immediately straightforward, but I think I've worked out how to create a clone of the LUNs. However, I'm not sure I know whether or not those LUNs are immediately brought up as read-writeable - either by performing a straight "LUN Clone" command or "Clone from snapshot" command.

    Anyone know?

    Like I said, its a long shot, but a lot of peeps on here have exposure to a hell of a wide range of technologies and I just wondered whether anyone had used NetApps in anger before

    Cheers!
     
    Certifications: A few
    WIP: None - f*** 'em
  2. zebulebu

    zebulebu Terabyte Poster

    3,748
    330
    187
    Out of the blue this, but I just thought that I'd post a follow-up to this, since I know people often trawl through old posts on this forum after Googling for answers!

    What I did in the end was:

    Disconnect from iSCSI targets
    Unmap all the existing LUNs
    Delete all the existing snapshots
    Create new snapshots of each LUN
    Create LUN clones from snapshots
    Make snapshots read-writable
    Re-map LUNs
    Add disk signatures on the server

    Pretty long-winded process all in all, but this is now the second time I've done it and I have the process pretty much set in stone now. The only thing that seems to present a problem, funnily enough, is not the NetApp stuff - the CLI is actually reasonably intutitive if a little cumbersome - but the disconnection from iSCSI targets & disconnecting the disks initially. This often fails for (seemingly) no reason at all, and is a bugger to try and recover from!

    Hope this helps anyone who a-Googling this way comes!
     
    Certifications: A few
    WIP: None - f*** 'em

Share This Page

Loading...
  1. This site uses cookies to help personalise content, tailor your experience and to keep you logged in if you register.
    By continuing to use this site, you are consenting to our use of cookies.