Skynet Skynet and Network Speed Test slow-downs ..

[capncybo]

BINGO!!!! Thanks & that makes perfect Sense.

So... Trend Micro is First in line & CPU intensive... a double whammy !!!
And I have tried withdrawing... (But didn't notice a great improvement)
However I missed... TRYING THE REBOOT AFTERWARDS.

Now... I really wanna Ditch work, go home & try again.
THX

 

[JaimeZX]

TrendMicro is sort of like an antivirus for your network. Have you ever had a computer not run slower with an AV client running?

Those of us with moderate internet speeds aren't going to see an impact, but if you have wicked fast speeds then apparently you might.

 

[capncybo]

This is No BIGGIE as my Client Run Tests give me numbers as 320/16 (On average)

But I wanted to update this thread in-case anyone was following along.
My ISP= 300/15 My Router= Asus RT-AC68U
As is mentioned clearly within the Asuswrt merlin notes...
-the Built-in Ookla Speed test will not be accurate @ Higher ISP speeds with lower powered routers.
Fine & Understood.

Except I have (read & been told) that using Skynet will NOT effect the results of the Built in Ookla Speed test.
But my actually testing seems to show otherwise.

When Opting out of Trend Micro & Disabling Skynet...
My Average Download speed with the Built-in Ookla Speed test is:
320Mbps (Averaged over 3-tests)

But when Opting out of Trend Micro & Enabling Skynet...
My Average Download speed with the Built-in Ookla Speed test was:
278Mbps (Averaged over 3-tests)

I have postulated the limiting factor may be:
The slow I/O of the router's USB controller accessing the swap-file, or maybe the USB-drive of which my Entware & Skynet is installed on.
And as others have mentioned... the swap drive does not appear to be heavily used.
But regardless, the numbers I posted "Above" are still my findings.

Perhaps it all breaks down to the limitations of the CPU & hitting ISP speeds above +275Mbps.

But to say that Skynet will not affect the Built-in Ookla Test does seem to be a FALSE Statement.
At least in my case.

 

[JaimeZX]

More interesting would be to run your experiment again on a hard-wired client device. I'm concerned the built-in speed test is too distracting to people at this point.

 

[capncybo]

More interesting would be to run your experiment again on a hard-wired client device. I'm concerned the built-in speed test is too distracting to people at this point.

My Guess is... The Numbers will Not Change when skynet is Enabled Or Disabled.

As said above, my fastest WiFi Client can already get 320/16 (On average) with Skynet-On & TrendMicro-Enabled.
Why would a 1G Wired Client come in @less?
Unlesss the Client CPU was too slow & couldn't handle a web-based Ookla SpeedTest up-to 330Mbps.

 

[capncybo]

Simply-Put, when testing with speeds >150Mbps... One of My Router CPU's is always MAXED out.

I Found this elsewhere...
If you are running TCP speed tests it has a lot to do on your CPU.
1: Open up task manager, have a look at your CPU (I SSH'd into the router & started HTOP)
2: Open up Chrome, have a look at your CPU. (I Logged into the router & went to the speed-test)
3: Run a speed test and see the CPU rise.
If you have an old CPU and its maxed out at 100% the test would not be able to show you the full speed you could receive. (One CPU is always Maxed-out when downloading)
Run on a new CPU and it will show better numbers.
Again this is just when you run speed tests, the internet would always be the same speed as you receive it from the ISP, it just means some computers can not handle the JAVA, Flash downloads.