CHR 7.0 Beta 5 CPU load on one core goes to 100%

I am running CHR version 7.0 Beta 3 that I upgraded from Version 6.46beta55. I have noticed that after a fresh boot one of the CPU cores will go to 100% usage. The maxed out core will swap around from Core0-Core3 but it stays at 100% usage. If I try and run the profiler I get the error Couldn’t start. My config is attached as a txt file. It was too long to paste in the forum.

Looking at the logs as its running it appears to be when my firewall adds a entry to the address list for SSH blacklist. When it added the entry a few moments ago the cpu went back up to 100% on core 1. This was after a fresh reboot.
Config.txt (32.9 KB)

Try moving your Layer-7 filter rule at bottom of chain, and do some optimization of firewall filter rules. Looks like your firewall is culprit here.
It would help a lot to track issue if you post screenshot of tools->profile.

I had the same issue. After update from stable to v7.0b3 the CPU load on my RB951G-2HnD were hitting 100% without any reason. Downgrade back to stable version fixed the issue.

tools->profile does not work in v7b3

FYI this firewall worked without issue on version 6. It is a modified version of Rick Frey’s firewall. The only thing I added was the IOT stuff and the layer7 rule which has always been at the top, and much of the time is disabled.


Sent from my iPhone using Tapatalk

I disabled the rule that 3 stage ssh rule and the cpu has stayed at normal usage. I believe the process that is causing the issue is the logging. I have re-enabled the ssh rule but disabled the logging. Will see how it runs with this.

Update The CPU still went up to 100% with the SSH rule enabled with logging disabled. I am now unsure what is causing it. It could be when it adds the IP address to the access list but not really sure.

This is still not fixed in 7.0 Beta 4. I re-enabled the SSH 3 stage firewall rules and again after a few minutes the CPU usage spikes. Just to be clear there is no attack happening on my router just normal port 22 stuff that I am sure everyone sees when that is exposed. I disabled the SSH rules again and re-booted and the CPU is fine. The process that is pegging the CPU is the routing process. The only way to fix it is a reboot. It moves from core to core but same issue of one core pegged at 100% Also there is less than 1Mbps of traffic going through my router right now and the issue persists.

Here is a snip of the profile tool with my current inbound and outbound traffic. As I stated previously the only way to fix this is a reboot.
Capture.PNG

This is still not fixed in 7.0 Beta 5. This worked fine for years on 6.x firmware.

7.0 Beta 8 seems to have fixed this issue. I have been running for over 2 days with multiple hits to my SSH acl and the cpu remains under control.