I'm having an issue with my 11.10 install of MythBuntu that lives on an ESXi 4.1u2 box. The MythBuntu Backend guest has 2 CPUs assigned to it, with 6GB of RAM. The recording storage lives on another VM (FreeNAS) on the same host (not using VMXNET3 driver, just Intel 1GB) that's accessed via NFS. I'm seeing this in 'top':<div>
<br></div><div><div>joltman@mythbackend:/var/log/mythtv$ top</div><div>top - 21:29:39 up 43 days, 9:14, 1 user, load average: 17.58, 17.65, 18.64</div><div>Tasks: 139 total, 2 running, 137 sleeping, 0 stopped, 0 zombie</div>
<div>Cpu(s): 2.6%us, 12.8%sy, 30.8%ni, 0.0%id, 0.0%wa, 2.6%hi, 51.3%si, 0.0%st</div><div>Mem: 6125956k total, 5282156k used, 843800k free, 65188k buffers</div><div>Swap: 0k total, 0k used, 0k free, 4511196k cached</div>
<div><br></div><div> PID USER PR NI VIRT RES SHR S %CPU %MEM TIME+ COMMAND</div><div> 9 root 20 0 0 0 0 S 76 0.0 37424:33 ksoftirqd/1</div><div>25613 mythtv 20 0 639m 29m 19m S 64 0.5 6:31.08 mythbackend</div>
<div> 3 root 20 0 0 0 0 S 49 0.0 34413:04 ksoftirqd/0</div><div>25647 mythtv 37 17 523m 209m 19m R 9 3.5 1:15.52 mythcommflag</div><div>25093 root 20 0 0 0 0 S 1 0.0 0:59.75 kworker/0:1</div>
<div> 2042 joltman 20 0 201m 13m 4564 S 0 0.2 154:27.87 xfce4-panel</div><div>25658 joltman 20 0 21468 1360 992 R 0 0.0 0:00.45 top</div><div> 1 root 20 0 24300 2216 1192 S 0 0.0 0:02.70 init</div>
<div> 2 root 20 0 0 0 0 S 0 0.0 0:01.15 kthreadd</div><div> 6 root RT 0 0 0 0 S 0 0.0 0:00.00 migration/0</div><div> 7 root RT 0 0 0 0 S 0 0.0 0:00.00 migration/1</div>
<div> 11 root 0 -20 0 0 0 S 0 0.0 0:00.00 cpuset</div><div> 12 root 0 -20 0 0 0 S 0 0.0 0:00.00 khelper</div><div> 13 root 0 -20 0 0 0 S 0 0.0 0:00.00 netns</div>
<div> 15 root 20 0 0 0 0 S 0 0.0 0:06.33 sync_supers</div><div> 16 root 20 0 0 0 0 S 0 0.0 0:00.97 bdi-default</div><div> 17 root 0 -20 0 0 0 S 0 0.0 0:00.00 kintegrityd</div>
<div>joltman@mythbackend:/var/log/mythtv$</div></div><div><br></div><div>Why is the ksoftirqd using 76% cpu? MythBackend process is very high. Even when I'm not playing or recording anything. I've done a cat /proc/interrupts and this stood out:</div>
<div><br></div><div><div> 18: 588914093 0 IO-APIC-fasteoi ioc0, eth0</div></div><div><br></div><div>I know the first thing people will suggest will be to move to 0.25. I'm actually thinking of jumping straight to 0.26 beta. Does anyone have any advice or things I should look for? Thanks!</div>