CPU Core 2 100% in 384.6

[Kal1975]

I'm using 384.6 on an RT-AC3100. Something interesting happened the other day. Everything seemed to be working fine. When I went to the router GUI, for some reason, I noticed that Core 2 was at 100%. When I checked TOP through ssh, there didn't seem to be any process using a corresponding amount of CPU resources. I remembered reading that restarting HTTPD might fix this. I couldn't figure out or find out the correct command to do this. So, I rebooted the router through the GUI. After the GUI became responsive again, the CPU for core 2 was still at 100%. Then I turned the router off, left it off for 1/2 to 1 minute. After that, the CPU for core 2 was back to normal at 0-3% or so.

Another interesting thing that happened was that the router could not reconnect to the internet until after I turned it off and then back on and I think it took a little while to reconnect as well. I'm thinking that it may not be related to the 100% issue and just may be a coincidence.

I also noticed that a day or so before this happened, there was a Two-way IPS detected. I was wondering if the 100% was related to this.

Last thing I did to see where the 100% was coming from was to look at the results of http://192.168.1.1/cpu_ram_status.asp?_=### It looks like this URL is what provides the GUI with the data to determine the CPU % and ram usage. The memory info seems obvious in the results. However, the CPUInfo numbers kept going up and were not obvious. Here's a sample of the output

var cpuInfo, memInfo = new Object(); cpuInfo = {"cpu0":{"total":"13071040","usage":"350289"},"cpu1":{"total":"13071002","usage":"294963"}} ; memInfo = {"total":"524288","free":"322552","used":"201736"};

So, can someone comment and explain how the % is determined from the data returned and does anyone have any ideas about the cause of the 100% CPU core usage?

 

[quant88]

There were similar posts previously. You may want to install htop to further observe which process is taking up the affected Cpu.

opkg install htop

 

[geobernd]

I had the same thing happen yesterday morning on my AC-68U. I noticed a little sluggishness on my network and when I checked the router I saw one core (don't remember if it was the first or second) at 100%. Looking at top it showed CPU 50% NIC .
I then decided to just witch to the 384.7 Alpha so sorry I don't have logs etc to share.

What I noticed later is that my ISP gave me a new WAN IP very recently (maybe within the last couple of hours before the issue as my network monitor gadget on the main PC still showed the old one but the external IP on the router main page was updated) - that usually doesn't happen more than once a year or so and I wonder if maybe the IP change caused a hickup...

I don't use VPNs or anything like that but I have DDNS setup and I use QOS with FreshJR Adaptive Script and have AIProtection on...

 

[NickBurns]

Had CPU1 100% IO wait apparently since August 29th tied to the last time syslog had written to the /jffs/syslog.log. Used "ps|grep D" to see what processes aren't sleeping well (STAT=D). For me it was the "touch /log/syslog". Then backtracked to note via GUI that System Log -> General Log had no output... killed process. Recovered.

 

[don_bidyre]

I just had the same issue with my AC-68U.
I performed a soft reboot & the 100% CPU (core1) remained. I then noticed that the router was reporting a problem with my ISPs DHCP & internet access was down (unfortunately I can’t say for certain if this was the case before the soft reboot, but I didn’t notice a DHCP/internet issue beforehand).
Another soft reboots & no change, but a hard reboot fixed it.
Another point that may be relevant is that before the soft reboot, I noticed that a few wireless devices that have been out of the house for a few weeks we’re still showing as online (& reporting a reasonable signal quality) within the device map (some on each of the 2 bands). These devices remained in this status through the soft reboots, but corrected after the hard reboot.
Obviously I’m assuming the soft reboots actually did reboot & not just display the progress percentage & carry on running the same SW instance regardless

 

[don_bidyre]

Added info, just looked at my system log & it seems nothing was logged since 5th Sept until my hard reboot (previous hard reboot was still in the log from 2nd Sept, so I only got 3 days worth before logging stopped .

 

[NickBurns]

So it sounds like you had the same issue -- syslog had stopped (for whatever reason) and couldn't write to the file. For me it was not partition / disk full -- plenty of space. I never rebooted my router -- I killed the process / zero'd the file and recovered that way.