In a previous post I went over the basics of top. In this post I’m going over some more advanced features that can be used to diagnose problems.
Top will by default refresh every 3 seconds. If this is too often for you, you can specify how many seconds to wait between each refresh:
1
top -d 10
d
stands for delay.
Another useful option is to hide idle tasks:
1
top -i
The letter i
can also by used to toggle idle tasks while top is running.
If you are interested in processes belonging to a specific user:
1
top -u adrian
Or type the letter u
while top is running. A prompt will appear to enter a user:
Another userful option is V
. It can be used to see the processes as a tree (see which process is a child of which):
Sometimes besides seeing the processing running on your system you might want to kill them. You can use k
to kill a process. You will be prompted for the PID and signal:
Working with columns
Another useful tool for finding what you are looking for is to work with the columns. While on top, use x
to turn on column highlighting. This shows the currently selected column:
You can move between columns using <
and >
for left and right respectively. The task list will be sorted by the selected column. You can invert the sorting by using R
. There are some bookmarks you can use to navigate between columns faster:
M
- %MEMN
- PIDP
- %CPU
You can search for a string in the highlighted column using L
:
You can also filter the tasks based on a column. To start filtering enter o
and enter a filter. An example filter can be: COMMAND=fire
, which will only show commands containing the word fire. You can also find the processes using more than certain amount of memory: %MEM>1.0
This are the parts I found most interesting about top. If you know any trick I didn’t cover, please let me know.
debugging
linux
]