The CPU utilization has reached ??%, which exceeds the pre-defined value of 80%.

[tekguru]

Occasionally (well once a day or so) I see:

SYSTEM ResourceMonitor. The CPU utilization has reached 93%, which exceeds the pre-defined value of 80%.

In the system log (overview). It shows as varying percentages; 83%, 93%.

Any ideas as to what can be causing this? I had thought the NAS needed a clean out (which it has now had yesterday) but I still got the warning last night. I've looked in the CPU section of the resource monitor but can't even see it spiking that high.

 

[fredbert]

Do you know what it was doing? Timed tasks, transcoding media, updating AV sigs.

 

[Telos]

SYSTEM ResourceMonitor. The CPU utilization has reached 93%, which exceeds the pre-defined value of 80%.

That must be new to DSM7. I see no such notification under DSM6.

That said, CPUs routinely burst to near 100%, so I'm unsure how that is handled. On my "widdle 213j", the CPU is always pegged above 97% when performing extended transfers, parity checks... Nearly every operation sends it above 80% for a while.

I don't understand the 80% CPU threshold for your system. Surely a more astute member will speak out on this.

 

[tekguru]

Nothing special that I'm aware of. Possibly Synology 2 backup or one of the events. These are the only events that take the CPU that high. The 80% threshold I thnk was set in DSM7 (not sure though).

 

[fredbert]

You could add a load of performance alarm rules for the Services so that they are likely to be triggered along with the System 80% CPU rule, e.g.

Or increase the System CPU rule to 90% or 95%.

 

[Rusty]

If you can catch that alarm and log into nas via ssh then you can run htop and see what process is top

 

[Telos]

Seems like a nuisance alarm, unless it's paired with a duration. Even so, you would want the ability to add exclusions (or possibly inclusions) to certain tasks.

 

[tekguru]

The problem so far is that it only seems to trigger in the middle of the night when I'm not around... I'll see if I can set a few service rules and see what that does.

 

[tekguru]

Okay I've set a warning for all services when they go past 60% CPU usage. We'll see what triggers.

 

[tekguru]

Well.. huh! I set up warnings for all services (as mentioned above) and not one triggered even though I got a 83% warning overnight. So I guess it is something I'll have to live with....

 

[tekguru]

Well since removing the DHCP Server package I've not suffered one CPU utilisation issue warning at all. Hoping it stays that way!