Netgate Discussion Forum
    • Categories
    • Recent
    • Tags
    • Popular
    • Users
    • Search
    • Register
    • Login

    PFsense SLow and High CPU

    Scheduled Pinned Locked Moved General pfSense Questions
    23 Posts 6 Posters 5.2k Views
    Loading More Posts
    • Oldest to Newest
    • Newest to Oldest
    • Most Votes
    Reply
    • Reply as topic
    Log in to reply
    This topic has been deleted. Only users with topic management privileges can see it.
    • H Offline
      heper
      last edited by

      also might try

      sysctl dev.em.0.debug=1
      ```from console
      1 Reply Last reply Reply Quote 0
      • D Offline
        doktornotor Banned
        last edited by

        karlzre - Apparently not in your case when idiots/broken mobile shit keeps hammering the CP over and over again. (Other than that, you told us exactly NOTHING about your "platform").  And yeah, there's also something very broken about the NICs. Did you enable some nonsense like polling there?

        1 Reply Last reply Reply Quote 0
        • K Offline
          karlzre
          last edited by

          also might try
          Code: [Select]
          sysctl dev.em.0.debug=1
          from console

          dev.em.0.debug: -1 -> -1
          
          
          Intel(R) Atom(TM) CPU N270 @ 1.60GHz
          Current: 800 MHz, Max: 1600 MHz
          2 CPUs: 1 package(s) x 1 core(s) x 2 HTT threads
          

          Memory usage
          13% of 2006 MB
          SWAP usage
          0% of 4096 MB
          Disk usage
          / (ufs): 2% of 23G
          /var/run (ufs in RAM): 3% of 3.4M

          1 Reply Last reply Reply Quote 0
          • K Offline
            karlzre
            last edited by

            Did you enable some nonsense like polling there?

            Enable device polling is disabled.

            1 Reply Last reply Reply Quote 0
            • D Offline
              doktornotor Banned
              last edited by

              @karlzre:

              Intel(R) Atom(TM) CPU N270 @ 1.60GHz
              Current: 800 MHz, Max: 1600 MHz
              2 CPUs: 1 package(s) x 1 core(s) x 2 HTT threads
              

              This CPU is just old slow POS, but that thing is running at half the freq it's spec'ed for with 100% CPU usage?  :o Try fiddling with the powerd settings.

              Also, post some information about the NICs.

              1 Reply Last reply Reply Quote 0
              • S Offline
                SplineZ
                last edited by

                @doktornotor:

                This CPU is just old slow POS, but that thing is running at half the freq it's spec'ed for with 100% CPU usage?  :o Try fiddling with the powerd settings.

                Also, post some information about the NICs.

                I have a bit more power behind my box and Ive run into the same thing.. Is powerd on minimum capable of causing this issue?

                my machine gets loaded up with 60mbits of torrent/netflix traffic (that is my connection speed) from time to time and I see that [kernel{em0 taskq}] and [intr{swi5: fast taskq}] tasks taking up the majority of the available CPU.. Polling is turned off..

                This machine as built from spare parts that I had laying around is it possible that im just running into a limitation with the hardware?

                IC2D 2.13ghz, Asus OEM shitty mobo, 3gb ram, Intel Pro/1000 Legacy (external), Realtek 8169 (Internal)

                1 Reply Last reply Reply Quote 0
                • D Offline
                  doktornotor Banned
                  last edited by

                  @SplineZ:

                  I have a bit more power behind my box and Ive run into the same thing.. Is powerd on minimum capable of causing this issue?

                  You need to fiddle with whatever settings are available there. People here mentioned boxes where CPU is never run at full freq unless powerd is enabled somehow. Sure if set @minimum, that's a good candidate.

                  1 Reply Last reply Reply Quote 0
                  • S Offline
                    Stewart
                    last edited by

                    You may also want to look at https://forum.pfsense.org/index.php?topic=59028.0 to remove those messages from the log and help cut down the noise.  Looks like they may not mean anything.

                    1 Reply Last reply Reply Quote 0
                    • S Offline
                      SplineZ
                      last edited by

                      @doktornotor:

                      @SplineZ:

                      I have a bit more power behind my box and Ive run into the same thing.. Is powerd on minimum capable of causing this issue?

                      You need to fiddle with whatever settings are available there. People here mentioned boxes where CPU is never run at full freq unless powerd is enabled somehow. Sure if set @minimum, that's a good candidate.

                      Interestingly enough, I switched it to adaptive and the overall load on the machine has lowered a HUGE amount. Lets say from 6.0 to under 0.7-0.5..  Kind of ironic that the feature meant to save energy is actually causing an increase… (in this situation)

                      Seriously, Turn on the profile "minimum" and max out your connection, your load will raise slowly.. turn it to Adaptive and watch it lower almost immediately. "[kernel{em0 taskq}]" is the task using all the cpu..

                      
                      With "Minimum" power setting on powerd
                      
                      last pid: 11607;  load averages:  3.36,  1.79,  0.99 
                        PID USERNAME      PRI NICE   SIZE    RES STATE   C   TIME    WCPU COMMAND
                          0 root          -92    0     0K   192K CPU0    0 647:46  79.49% [kernel{em0 taskq}]
                         12 root          -56    -     0K   272K WAIT    1 432:38  51.46% [intr{swi5: fast taskq}]
                      87987 root          100   20 25160K  6464K CPU0    0   0:04  26.27% /sbin/pfctl -ss
                         11 root          155 ki31     0K    32K RUN     1 390.8H  15.97% [idle{idle: cpu1}]
                         11 root          155 ki31     0K    32K RUN     0 385.3H  11.18% [idle{idle: cpu0}]
                      
                      With "HiAdaptive" power setting on powerd
                      
                      last pid: 61673;  load averages:  0.53,  1.23,  0.89   
                           PID USERNAME      PRI NICE   SIZE    RES STATE   C   TIME    WCPU COMMAND
                         11 root          155 ki31     0K    32K RUN     1 390.9H  97.56% [idle{idle: cpu1}]
                         11 root          155 ki31     0K    32K CPU0    0 385.3H  83.98% [idle{idle: cpu0}]
                          0 root          -92    0     0K   192K -       0 648:19  10.16% [kernel{em0 taskq}]
                         12 root          -56    -     0K   272K WAIT    1 433:00   5.66% [intr{swi5: fast taskq}]
                         12 root          -92    -     0K   272K WAIT    1  58:56   0.98% [intr{irq19: em0 uhci1}]
                         12 root          -92    -     0K   272K WAIT    0  40:54   0.78% [intr{irq16: re0 uhci3}]
                      
                      

                      No other changes were made on the machine other then fiddling with the powerd settings..  I never would have thought to look there.. Kudos and thanks!

                      1 Reply Last reply Reply Quote 0
                      • H Offline
                        Harvy66
                        last edited by

                        You're misunderstanding what're being reported. Load is relative. If your CPU is running at 100mhz, then it's not hard to reach 75% usage. The reason "load" is high when using minimum is because the CPU is being down-clocked, you are saving power.

                        1 Reply Last reply Reply Quote 0
                        • First post
                          Last post
                        Copyright 2025 Rubicon Communications LLC (Netgate). All rights reserved.