I would suggest you log into NOW, and search for "CPU is high on my storage system" .

CPU spikes are normal, and as the NetApp guy mentions, it's likely just the asup daemon going off. Have you tried triggering a manual autosupport and seeing if the DFM alert goes off? It's also worth noting that scrubs kick off around this time, and can contribute.

On a single processor machine (or even dual, if I'm remembering the 2050 right), there aren't a lot of CPUs to share the load.

But again, see first line. DOT uses CPU all the time, very efficiently, and it isn't a good measure of your storage system being overloaded. If your DFM install can support multi...point(wrong word)? triggers (i.e., CPU high *and* LATENCY > X), use that. Or, just use latency for protocol. That's how performance is usually impacted.

--- On Mon, 8/17/09, Halton, Robert rhalton@tpg.com wrote:

From: Halton, Robert rhalton@tpg.com Subject: RE: 2050 with high CPU Usage during weekly asup To: toasters@mathworks.com Date: Monday, August 17, 2009, 12:57 PM

Lets say for argument sake that the system is a brand new install with nothing on it apart from the root vols.

And that I have 3 of them..   And they all do the same thing.

Was just looking for feedback from the world at large before opening my ticket.

From: Jeff Mohler [mailto:speedtoys.racing@gmail.com]

Sent: Monday, August 17, 2009 8:06 AM

To: Borders, Rich

Cc: Halton, Robert; toasters@mathworks.com

Subject: Re: 2050 with high CPU Usage during weekly asup

Or it could be nothing at all.

High CPU doesnt equal high latencies or slow user performance by itself.

Backups will bump up CPU, on slower systems more than faster ones..capture more data.

On Mon, Aug 17, 2009 at 7:44 AM, Borders, Rich Rich.Borders@netapp.com wrote:

Well, the asup is a list of commands that generate a lot of file responses that are then later sent off to the asup tool NetApp uses. It could be that it puts stress on vol0 (or root vol) and that is generating the load. (unless you can verify that DFM is erroneous)

Rich

From: Halton, Robert [mailto:rhalton@tpg.com]

Sent: Monday, August 17, 2009 10:30 AM

To: toasters@mathworks.com

Subject: 2050 with high CPU Usage during weekly asup

Has anyone else seen this?

A 2050 Cluster when it sends the weekly ASUP will max out CPU, or at least DFM tells me so..

*** Event details follow.***

General Information

---

DataFabric Manager Serial Number: 1-50-008777 Alarm Identifier: 4

Event Fields

---

Event Identifier: 2327284

Event Name: CPU Too Busy

Event Description: CPU utilization on an appliance Event Severity: Warning Event Timestamp: 16 Aug 03:21

Source of Event

---

Source Identifier: 4111

Source Name: RCNETAPP2 Type: Active/Active Controller Source Status: Warning

Event Arguments

---

cpuTooBusyThreshold: 95

cpuUtilization: 97.6871

--NetApp DataFabric Manager

This message is intended only for the person(s) to which it is addressed and may contain privileged, confidential and/or insider information. If you have received this communication in error, please notify us immediately by replying to the message and deleting it from your computer. Any disclosure, copying, distribution, or the taking of any action concerningthe contents of this message and any attachment(s) by anyone other than the named recipient(s) is strictly prohibited.

--

No Signature Required

Save The Bits, Save The World!

This message is intended only for the person(s) to which it is addressed and may contain privileged, confidential and/or insider information. If you have received this communication in error, please notify us immediately by replying to the message and deleting it from your computer. Any disclosure, copying, distribution, or the taking of any action concerning the contents of this message and any attachment(s) by anyone other than the named recipient(s) is strictly prohibited.