Tags
I believe this feature has been around since the DFM v4.x releases. If you haven’t noticed, DFM has been re-branded to NetApp OnCommand Unified Manager. There are Core and Host packages. You can download the latest v5.2 release here. You need a support contract to access the site.
Note- click images in the post below to increase size.
The NetApp OnCommand Unified Manager Core package bundles a great performance analysis tool built into the NetApp Management Console. Once you install OnCommand Unified Manager Core, its available in two locations:
1) C:\Program Files\NetApp\DataFabric Manager\DFM\web\clients\nmconsole-setup-3-3-win32.exe.
2) Start > NetApp > DataFabric Manager > Show Appliance Summary Page > click Setup > click Download Management Console.
2a) Click Download Windows Installation (version 3.3) and save it locally.
Install the Management Console on your workstation or server.
1) Double-click nmconsole-setup-3-3-win32.exe > Next > Install > Next > Finish. It will launch when done.
2) Enter your OnCommand Unified Manager server name or IP, username and password and click connect. Note you can click Options and switch between http and https. Hopefully you are using https.
3) In the upper left, click Tasks > Manage Performance. Under View, select Logical and click the controller that you want to assess.
4) In a few seconds the page will render. On the right, just above the Network Throughput diagnostic panel, click View Actions > Diagnostics.
5) You will see either green, yellow or red as indicators of the health check category. Click on each and see what is available. In this case, when you select NAS Specific Issues, it says that Atime updates are enabled on the volumes. See the recommendation at the bottom and correct as needed.
Note: You can sometimes improve performance by directing Data ONTAP to skip logging of access time (atime) information to NVRAM. The downside is that if there is a storage system crash, a few seconds worth of access time updates may not be recorded in the file system.
To make the change:
> vol options <your_vol_name> no_atime_update on
6) In this image for SAN Specific Issues, you can see it has detected LUN partial read/write issues. Assess the recommendations and make changes as needed.
Note: notice on the left that you can adjust the date and time of the diagnostic. This is useful when you want to assess a change you made to an NFS mount, for example rsize=8192,wsize=8192, or maybe a realigned LUN, etc. You can go back in time and correlate the diagnostic and performance data.
7) See TR-4090 (page 46) for Diagnostic tests and meanings.