Just found reason for high load average


#1

I just noticed that shortly after start, load average is always 1.0 or more.

This does not make sense, becouse nothing serious is actually running.

Command ‘top’ confirmed that.

Then  digged more in google anf found this command:

top -b -n 1 | awk '{if (NR <=7) print; else if ($8 == "D") {print; count++} } END {print "Total status D: "count}'

And result of this command on My Book Live is following:

top - 20:12:53 up 1 day, 57 min, 3 users, load average: 1.01, 1.00, 1.00
Tasks: 64 total, 1 running, 63 sleeping, 0 stopped, 0 zombie
Cpu(s): 2.2%us, 0.9%sy, 0.2%ni, 96.0%id, 0.7%wa, 0.0%hi, 0.1%si, 0.0%st
Mem: 254336k total, 224192k used, 30144k free, 58880k buffers
Swap: 499584k total, 17600k used, 481984k free, 98176k cached

  PID USER PR NI VIRT RES SHR S %CPU %MEM TIME+ COMMAND                                                                                                                                                                                            
  375 root -2 0 0 0 0 D 0.0 0.0 0:00.00 btn_t                                                                                                                                                                                              
Total status D: 1
MyBookLive:~#

 So… that mystical btn_t process is reason for this ‘always 1.00’ load average…

I have no idea what it is and how can i disable it from running.

Ideas?


#2

Seriously, I’m just curious…

Why do you care?   The performance of this little box is quite good as it is.   Are you just looking to eek out another 2 mbit/second or something?   :smileyvery-happy:

Why do you want to disable it? 


#3

It just feels wrong… It is not about getting 2mbit/s more.

In MyBook World typical idle average is 0.00 or 0.01. I expect same for Live.

Also i’m suspect (maybe i’m wrong?) it might heat up hdd temp which is actually quite hot (even seen 56C).

I digged more and found that this:

MyBookLive:/proc/375# lsof -p 375    
COMMAND PID USER FD TYPE DEVICE SIZE NODE NAME
btn_t 375 root cwd DIR 9,1 4096 2 /
btn_t 375 root rtd DIR 9,1 4096 2 /
btn_t 375 root txt unknown /proc/375/exe
MyBookLive:/proc/375#

 And file listing shows this:

MyBookLive:/proc/375# ls -l /proc/375
ls: cannot read symbolic link /proc/375/exe: No such file or directory
total 0
dr-xr-xr-x 2 root root 0 Jan 22 21:18 attr
-r-------- 1 root root 0 Jan 22 21:18 auxv
--w------- 1 root root 0 Jan 22 21:18 clear_refs
-r--r--r-- 1 root root 0 Jan 22 21:27 cmdline
-rw-r--r-- 1 root root 0 Jan 22 21:18 coredump_filter
lrwxrwxrwx 1 root root 0 Jan 22 21:27 cwd -> /
-r-------- 1 root root 0 Jan 22 21:18 environ
lrwxrwxrwx 1 root root 0 Jan 22 21:27 exe
dr-x------ 2 root root 0 Jan 22 21:27 fd
dr-x------ 2 root root 0 Jan 22 21:18 fdinfo
-r-------- 1 root root 0 Jan 22 21:18 limits
-r--r--r-- 1 root root 0 Jan 22 21:27 maps
-rw------- 1 root root 0 Jan 22 21:18 mem
-r--r--r-- 1 root root 0 Jan 22 21:18 mountinfo
-r--r--r-- 1 root root 0 Jan 22 21:18 mounts
-r-------- 1 root root 0 Jan 22 21:18 mountstats
dr-xr-xr-x 7 root root 0 Jan 22 21:18 net
-rw-r--r-- 1 root root 0 Jan 22 21:18 oom_adj
-r--r--r-- 1 root root 0 Jan 22 21:18 oom_score
-r-------- 1 root root 0 Jan 22 21:18 pagemap
-r-------- 1 root root 0 Jan 22 21:18 personality
lrwxrwxrwx 1 root root 0 Jan 22 21:27 root -> /
-rw-r--r-- 1 root root 0 Jan 22 21:18 sched
-r--r--r-- 1 root root 0 Jan 22 21:18 smaps
-r--r--r-- 1 root root 0 Jan 22 21:27 stat
-r--r--r-- 1 root root 0 Jan 22 21:27 statm
-r--r--r-- 1 root root 0 Jan 22 21:27 status
-r-------- 1 root root 0 Jan 22 21:18 syscall
dr-xr-xr-x 3 root root 0 Jan 22 21:18 task
-r--r--r-- 1 root root 0 Jan 22 21:18 wchan
MyBookLive:/proc/375#

Still not clue what the heck is it :frowning:


#4

It is the software polling the reset Button. The button is not tied to an interupt, so needs to be polled. The interval is pretty large, so should not be of any concern to overall system performance.

Tony


#5

Thanks for explanation!