USG Flex 200H - Strange CPU usage

GiuseppeR

Hello everyone,

as you can see with 1.21 firmware you have a strange CPU usage graph:

In my case I have about 25% on Core1 and 25% on Core2, while Core3 and Core4 are at zero value.

May I miss something somewhere else?

To be exact these are details for each CoreX:

PeterUK

I see over 50% on my Flex200H one far out idea is that its generated load so that when the device is at load it has time to handle the traffic?

GiuseppeR

It could be, but IMHO it seems strange. I would not expect a mathematical correspondence comparing each CoreX to the CPU general graph, but in this scenario the difference is huge while, for example, with RAM usage is accurate.

Also if I change the graph from 24 hours to real time the usage in % for each CoreX is far less than CPU usage.

Zyxel_Judy

Hi @GiuseppeR ,

In case, you want to compare the CPU usage graph's statis with the CPU information in the Maintenance menu, please select the "Current" option in the CPU usage graph.

However, it seems like you have another question, we'd like to clarify it. Please correct me if I've misunderstood anything.
You're asking about a discrepancy in the CPU usage reporting:

• In the CPU usage graph: Core1 = 25%, Core2 = 25%, Core3 = 0%, Core4 = 0%
• In the maintenance menu: CPU0 = 27.3%, CPU1 = 25.4%, CPU2 = 0%, CPU3 = 0%

Is your question about why the core numbering is different between these two displays (starting at Core1 vs. CPU0), right?

GiuseppeR

Hello @Zyxel_Judy

to be exact there are more questions.

I suppose that inside 200H there is 1 CPU with 4 Cores.

I think that there are not 4 different CPUs.

Honestly I do not unscrew the beta unit to check its board.

First question:

I see more than 60% of CPU usage in the dashboard.

Considering CPU made by 4 Cores and considering that only two of them are at 25% it is not mathematically possible that >60% is the sum of single usages for Cores.

Let's simplify it.

Core1: 25/100

Core2: 25/100

Core3: 0/100

Core4: 0/100

so if CPU is the sum of single Cores, I would expect a value of 50/400 that is 12,5%

So in the main graph:

the CPU value, in my opinion, should stay at about 12-13% instead of 62.2%

Second question:

If the CPU is only one, you should see Core1 + Core2 + Core3 + Core4.

Using computer based numbering your should see Core0 + Core1 + Core2 + Core3.

But this is only a formal appearance.

The problem I see is that on "CPU Usage" graph in the dashboard you see it splitted in:

"Core1 + Core2 + Core3 + Core4"

while on Maintenance area you see:

"CPU0 + CPU1 + CPU2 + CPU3"

Reading maintenance window gives you the idea that there are 4 CPUs instead of 4 Cores

I'm always at your disposal, best regards

Zyxel_Judy

Hi @GiuseppeR ,

Point 1: Current CPU status display:

• In Maintenance > Diagnostics > CPU/Memory Status > CPU: The value shown is the current CPU usage.
  
• On the Dashboard: The value displayed is an average of the last 60 seconds.

This difference in calculation methods allows users to view both current and average CPU usage, providing a more comprehensive overview of system performance.

Point 2: Display changes in the next version:

In Maintenance > Diagnostics > CPU/Memory Status > CPU: We will rename "CPU 0, CPU 1, CPU 2, CPU 3" to "Core 1, Core 2, Core 3, Core 4". This change will align the naming convention with the Dashboard's graph, make it easier for users to correlate information between different sections of the interface.