Disk Busy & Flash Cache Issue (FAS8040 ONTAP Release 9.1 & OnCommand Unified Manager 9.5)

[Anonymous]

Hello Dear NetApp Community,

We have several NetApp products and I stuck in a FAS8040 disk busy issue. Here is the thing,

On our NetApp 2240 Product (ONTAP 8.2), I can see the disk busy graph very clearly with NetApp Management Console 3.3.5 as you can see below. (View -> Physical -> Related Aggregate -> Metrics )

Disk Busy Line Chart On Management Console

Yes, this is very useful for me and I want to see the same chart for our other NetApp FAS8040 ONTAP Release 9.1 & OnCommand Unified Manager 9.5.

Is there any way to see this Disk Busy graph on FAS8040 ONTAP 9.1 or OCUM 9.5 ?

Also how can I know my 512GB Flash Cache on FAS8040 is enough for me? Is 512GB Flash Cache is enough or should i increase to 1TB Flash Cache?.  I want to know how much of these 512 gb Flash Cache drive is being used. Any help would be great.

Thanks in advance,

Mert

[colsen]

Hello,

I too miss the NMC graphs that just showed the "top" objects.  We'd even use that disk-busy graph to see if we had a disk in a pre-fail condition (since one disk would usually go off the rails when that happened).  Anyway, you can use the statistics command in ONTAP to gather some of that:

cluster-name::*> statistics disk show -node cluster-name-07

cluster-name : 3/19/2019 06:39:02

                     Busy *Total Read Write Read Write Latency
Disk  Node          (%)    Ops   Ops  Ops  (Bps) (Bps) (us)
----- -------- ---- ------ ---- ----- ------- -------- -------
7.0.5 cluster-name-07 2    1323 1112  211   8841216 11446272 290
...

It's not as user friendly as the NMC graphs were, but it's a start. 

On the FlashCache/PAM question, I'm going to have to dig a bit.  Predictive Cache Statistics (PCS) gets a little weird on a system that already has a FlashCache card installed.  That said, I used to peruse my top volumes on an aggregate and see what their cache hit ratio was per this page on the doc site:

http://docs.netapp.com/ocum-72/index.jsp?topic=%2Fcom.netapp.doc.onc-um-perf-ag%2FGUID-40A72B78-E550-4ACC-B823-C15162C7F0B8.html&resultof=%22%63%61%63...

That data doesn't seem to be forthcoming in our OCUM 7.3P1 tool.  If I can't shake something loose I'll ping some folks who might be able to track that down.

Regards,

Chris

[ryanjenson90]

hello,

i guess the solution given by @colsen works well.

try using that.

You can also try this guide on using vector software if you dont find solution after first thing.

Regards.

[xandervanegmond]

Hi Mert,

Not only would I advise getting used to the command line (qos is a very very powerful command to troubleshoot performance issues), you might also want to look into OnCommand Performance Manager.

NSM no longer contains disk busy graphs, but those are available in OPM.

/Xander