vFiler migrate chooses destination aggregate without asking

[fletch2007]

Hi,

vFiler migrate chooses the destination aggregate without asking – I’m guessing its using a simple algorithm which may choose the first aggr with enough free space?
What if we want the migrated vfiler to end up on a specific AGGR when there are multiple that have enough space?
(I guess I could fill the ones I don’t want it to choose with dummy volumes – but is there a better way?)

thanks

[rshiva]

Hi there,

You won't be able to do that through NMC, however you can achieve it through the CLI:

C:\Users\Administrator>dfpm migrate start help

NAME
    start -- start migration of a vFiler unit or dataset

SYNOPSIS
    dfpm migrate start [ -D ] -f <destination-storage-system> | <resource-pool> | vol-aggr-map
        [ -d description ] [ -v <provisioning-policy-name-or-id> ]
        [ -i <ipbindings> ] [ -t ] [ -o [ -p <script-path> ] ]
        [ -k <max_xfer_rate in kbps> ] [ -r ][ -s ]
        <dataset-name-or-id | vfiler-name-or-id>

DESCRIPTION
    Starts migration of a vFiler unit or dataset
    -f specifies either the resource pool or storage system
    to which the vFiler unit or dataset needs to be migrated.
    Another option is to specify the actual volume to aggregate
    mapping while migrating a vFiler unit. The format is
    "vfiler-volume1=destination-aggregate1,vfiler-volume2=destination-aggregate2,..."
    A destination aggregate should be specified for each of the
    vFiler unit volumes. All the destination aggregates should
    belong to the same storage system.
    -d specifies a description for migration operation
    If -D option is specified, only dry run results will be
    displayed. Migration will not be started.
    -v specifies provisioning policy to be used to provision
    destination storage during migration.
    -i is a comma separated list of IP Addresses and interface
    bindings to be added to the vFiler unit. Each IP binding
    needs to be of the format
    ip_address:interface_name:netmask[:vlan_id][:mtu_size][:partner_interface].
    If a vFiler unit is being migrated, this argument is mandatory.
    Specify -o option to perform one step migration.
    Specify -t, if migration cutover has to happen online.
    -r specifies, route mode that needs to be migrated during cutover.
    Valid route modes are 'static', 'persistent' and 'none'.
    By default 'static' routes will be migrated. This option is
    valid only for online migration.
    -s If this option is specified, then the migrated routes will not be persistent
    on destination storage system after cutover. By default, the migrated routes
    will be persistent on destination storage system.
    -k Specify the bandwidth throttle to be used for baseline
    and snapmirror updates. Default is 0 i.e no throttle.
    -p specifies the complete path to the script to be run in
    pre and post cutover mode during one step migration.

Hope that helps.

Thanks and regards

Shiva Raja

[arndt]

Recent D releases of OM 4.0 include NMC 3.0D1, which gives you the option (from the NMC vfiler migration wizard GUI) of specifying the destination aggregate for each volume being migrated.

Mike

[fletch2007]

I don't see "Recent D releases of OM 4.0" for download on NOW - this is the latest one I see - (from Feb 18, 2010)

First Customer Shipment Release	Release Date
DataFabric Manager 4.0 (includes Operations Manager, Protection Manager and Provisioning  Manager)	18-FEB-2010

thanks

[arndt · Accepted solution]

http://now.netapp.com/NOW/download/software/dfm_win/4.0D16/

Mike

[fletch2007]

thanks Mike - got 3.0D1 and the aggr specify option works like a charm

[rshiva]

Hi,

If you're using Linux: https://now.netapp.com/NOW/download/software/dfm_lin/4.0D16/. The Link presented by Mike is DFM installation file for Windows. Plus don't forget to install the 3.0D1 client.

Thanks and regards

Shiva Raja