Updated to 384.11 - Both CPU Cores Running @ 97%

[Poseidon]

Did a clean flash of 384.11 on my 86U router. Set up everything from scratch. Including thr scripts. However I'm noticing that the CPU cores (1 + 2) are very high. Never seen this with 384.10.

I've rebooted the router multiple times and even left it on overnight but the cores are still running at 96-98%.

Please advice.

Thanks

 

[skeal]

Did a clean flash of 384.11 on my 86U router. Set up everything from scratch. Including thr scripts. However I'm noticing that the CPU cores (1 + 2) are very high. Never seen this with 384.10.

I've rebooted the router multiple times and even left it on overnight but the cores are still running at 96-98%.

Please advice.

ThanksView attachment 17546

SSH into the router and run top or htop if you have entware installed. That will tell you what process is hogging the cores.

 

[kernol]

Did a clean flash of 384.11 on my 86U router. Set up everything from scratch. Including thr scripts. However I'm noticing that the CPU cores (1 + 2) are very high. Never seen this with 384.10.

I've rebooted the router multiple times and even left it on overnight but the cores are still running at 96-98%.

Please advice.

ThanksView attachment 17546

A little "intimidated" to respond ...

Poseidon was god of the sea, earthquakes, storms, and horses and is considered one of the most bad-tempered, moody and greedy Olympian gods. He was known to be vengeful when insulted. He is the son of Cronus and Rhea and was swallowed by his father along with Hades, Demeter, Hestia and Hera.

and I see from prior posts elsewhere that you have serious issues about "Privacy" - but can I please appeal to you to at least complete a signature for this forum [along the lines of mine if you wish - see below] - so that when you raise queries ... those members who leap in to help at least know something about the Router and add-ons you are deploying. As a "Senior Member" I assume you know where to go to complete it.

I've had an issue similar to yours which turned out to be a faulty USB drive containing Entware and various scripts.
The router was going bezonks trying to execute scripts which had been corrupted.

 

[Poseidon]

Rebooted the router and everything came back normal. Both cores @ 0-2%

I previously modified a script (Diversion) so I'm not sure if that was the cause for the spike.

Installed htop as well and everything looked normal there.

Also I assume that the RAM being around 85-90% is NORMAL?

 

[skeal]

Rebooted the router and everything came back normal. Both cores @ 0-2%

I previously modified a script (Diversion) so I'm not sure if that was the cause for the spike.

Installed htop as well and everything looked normal there.

Also I assume that the RAM being around 85-90% is NORMAL?

Yes, I highly recommend a swap file configured through AMTM.

 

[Poseidon]

Yes, I highly recommend a swap file configured through AMTM.

Already did (2GB) initially after installing amtm.

 

[Poseidon]

Just noticed that anytime I make a change through the web GUI and click "apply" - both the CPU cores jump to 95-100% usage. However upon reboot is back down to normal (0-3%).

Is this immediately supposed to happen when making setting changes and applying them in the web GUI - it is it mandatory to reboot the router after every settings adjustment because the CPU cores are spiking?

 

[EmeraldDeer]

Just noticed that anytime I make a change through the web GUI and click "apply" - both the CPU cores jump to 95-100% usage. However upon reboot is back down to normal (0-3%).

Is this immediately supposed to happen when making setting changes and applying them in the web GUI - it is it mandatory to reboot the router after every settings adjustment because the CPU cores are spiking?

• Of course not
• Find CPU hog process with top or htop
• If no process, take note of usr, sys or other percentages

 

[L&LD]

Did a clean flash of 384.11 on my 86U router. Set up everything from scratch. Including thr scripts. However I'm noticing that the CPU cores (1 + 2) are very high. Never seen this with 384.10.

I've rebooted the router multiple times and even left it on overnight but the cores are still running at 96-98%.

Please advice.

ThanksView attachment 17546

What is a 'clean flash' for you? What specific steps did you take?

 

[Poseidon]

What is a 'clean flash' for you? What specific steps did you take?

I followed your guides (m&m and amtm).

Just thinking... Could it be due to FreshJR QoS initiating when a settings change is applied? Doesn't that script spike cpu for like 5 minutes or so?

 

[L&LD]

I followed your guides (m&m and amtm).

Just thinking... Could it be sure to FreshJR QoS initiating when a settings change is applied? Doesn't that script spike cpu for like 5 minutes or so?

Depending on the router and network conditions at the time, it is at least 5 minutes. I simply wait for about 10 minutes to know that any changes/tests I'm doing will be under actual conditions that the router will be in and not some in-between state.

 

[WuTang LAN]

A simple 'htop' or 'top' will identify the process that is hogging the CPU, in most cases.

Diversion will most likely have been the cause for the CPU usage. Seeing as you had just made a change, Diversion then- naturally- has to account for it.

90% RAM usage is perfectly normal in a Linux environment. Seeing as you already have a swap file, you're all set.

 

[joe scian]

I had a wierd issue with diversion when adding an address to whitelist - - both cores went to 100% indefintely and my WAN would show disconnected. Another symptom - when I tried to go to my syslog messages web page -again 100% on both cores - it never made it to that web page and then I would lose my Web Page logon and even an service restart_httpd wouldnt bring back the web logon page. This was on Beta 2 384.11.
Even a dirty flash to a prior beta - ie beta 1 would not clear the condition above. The only thing that cleared all these symptoms was a restore to factory default and manual configuration from the top - this i did when release version of 384.11 was released and I havent had any issues since.
PS - a full reinstall of diversion still didnt clear the condition above prior to the factory default restore