Notes from: ‘Demo of Clustered ONTAP Volume Move Capability’

The following are a few notes taken from watching the excellent NetAppTV YouTube Video -
- by Christopher Madden

Lab Elements

Windows 2008 Host, Fibre Channel Switch(es), Clustered ONTAP (4 nodes)

Consoles Used

To watch ALUA in action and see the Optimized path update:
(Win. 2008) Server Manager > Storage > Data ONTAP DSM Manager > Virtual Disks

To see read/write bytes/sec and more:
(Win. 2008) Reliability and Performance Monitor

Running sio_ntap to general some load:
(Win 2008) Command Prompt

And an SSH client!

Running the Demo

Via command prompt >

To generate some load:

sio_ntap_win32.exe 90 100 32k 5g 600000 24 img.iso

Via CDOT cluster shell ::>

To get statistics for the fibre channel traffic on a node:

statistics show-periodic -object fcp_lif -instance vs1.cdot-01-fl -counter read_data|write data

statistics show-periodic -object fcp_lif -instance vs1.cdot-03-fl -counter read_data|write data

(We run this in two separate SSH console windows - once for the node that is the current home for the volume, and once for the node that will be the new home for the volume to be moved.)

To get statistics on the Fibre Channel LUN itself:

statistics show-periodic -object lun -instance /vol/win_t_drive/t_drive -counter read_data|write_data|avg_latency

To move the volume:
volume move start -vserver vs1 -volume win_t_drive -destination-aggregate aggr1_cdot03 -foreground true

(The ‘foreground true’ is optional but useful for watching the progress of the volume move - the cluster shell prompt becomes available again after the job has finished.)

Note: If you disable the Optimized path (which is on the node that homes the volume), traffic to the FC lun travels equally down the Non-optimized paths (i.e. the other nodes each take a share of the traffic to the Windows host, with the data from/to the volume traversing the cluster interconnect.)

Comments