Directadmin always at 100% cpu

jonium

Verified User
Joined
Nov 10, 2010
Messages
208
Location
Alezio - Lecce- Apulia - South Italy
Hello,
here is the related raw from the process monitor on DA (top):
11808 nobody 20 0 90664 54364 2496 R 100.0 0.1 231:34.46 /usr/local/directadmin/directadmin d

How can I watch what is it doing?
 
# lsof -p 11808
COMMAND PID USER FD TYPE DEVICE SIZE/OFF NODE NAME
directadm 11808 nobody cwd DIR 9,2 4096 93194993 /usr/local/directadmin
directadm 11808 nobody rtd DIR 9,2 4096 2 /
directadm 11808 nobody txt REG 9,2 8680072 93195462 /usr/local/directadmin/directadmin
directadm 11808 nobody mem REG 9,2 555912 93194096 /usr/lib64/libfreeblpriv3.so
directadm 11808 nobody mem REG 9,2 61624 93194453 /usr/lib64/libnss_files-2.17.so
directadm 11808 nobody mem REG 9,2 11392 93194247 /usr/lib64/libfreebl3.so
directadm 11808 nobody mem REG 9,2 2156160 93194067 /usr/lib64/libc-2.17.so
directadm 11808 nobody mem REG 9,2 88776 93195467 /usr/lib64/libgcc_s-4.8.5-20150702.so.1
directadm 11808 nobody mem REG 9,2 1137024 93194370 /usr/lib64/libm-2.17.so
directadm 11808 nobody mem REG 9,2 991616 93194043 /usr/lib64/libstdc++.so.6.0.19
directadm 11808 nobody mem REG 9,2 43776 93194456 /usr/lib64/librt-2.17.so
directadm 11808 nobody mem REG 9,2 142232 93194333 /usr/lib64/libpthread-2.17.so
directadm 11808 nobody mem REG 9,2 105824 93194455 /usr/lib64/libresolv-2.17.so
directadm 11808 nobody mem REG 9,2 19288 93194348 /usr/lib64/libdl-2.17.so
directadm 11808 nobody mem REG 9,2 90248 93194218 /usr/lib64/libz.so.1.2.7
directadm 11808 nobody mem REG 9,2 40664 93194246 /usr/lib64/libcrypt-2.17.so
directadm 11808 nobody mem REG 9,2 163400 93192764 /usr/lib64/ld-2.17.so
directadm 11808 nobody 0r CHR 1,3 0t0 1028 /dev/null
directadm 11808 nobody 1u CHR 1,3 0t0 1028 /dev/null
directadm 11808 nobody 2u CHR 1,3 0t0 1028 /dev/null
directadm 11808 nobody 3r REG 9,3 20389269 142868486 /home/tmp/quota-dump.tQYlflM7C8k0VJZoKqLAy03dashOoAzhYXQEGmLQv3QOLzCfnVqlVrM5j3SXCRCk.progetto
directadm 11808 nobody 4u IPv6 1097716 0t0 TCP serverIP:EtherNet/IP-1->hostxx-xxx-dynamic.xxx-xx-r.retail.telecomitalia.it:50149 (CLOSE_WAIT)

serverIP is the IP of the server
hostxx-xxx-dynamic.xxx-xx-r.retail.telecomitalia.it is my IP
 
You might also consider doing something like:
strace -cirtv -p 11808
or
strace -cirtv -o /tmp/datrace.log -p 11808

The first will spew lots of system calls to the screen and it may get lost, the second redirects the output to the file /tmp/datrace.log. This will give you a trace of all the low level system calls that directadmin is making as well as how long it is spending in each call. I suggest doing the second format and let it run for a minute or two of high CPU usage, then CTRL-C out of it. Then open the /tmp/datrace.log file and look for system calls that that are either in a loop (good reason for high CPU) or ones that have a high relative times. The ones with the high relative times may point you to a reason why the CPU usage is high. You may or may not want to add the 'f' flag to the above commands, that tells strace to also follow child processes, which might also be where the issue is. I would do it without the 'f' flag first so there are fewer lines to parse.
 
# cat /tmp/datrace.log
% time seconds usecs/call calls errors syscall
------ ----------- ----------- --------- --------- ----------------
100.00 0.000023 0 377 read
0.00 0.000000 0 38 brk
------ ----------- ----------- --------- --------- ----------------
100.00 0.000023 415 total
 
Back
Top