Diagnaosticsaur

Questions? Email: [email protected].

Scripts

• freespace.pl -- Easy parser for df -h. Usage:

  perl freespace.pl --filesys="/dev/sda1" --option="Size" > output

where each filesys is a drive and each option is a column on the df -h table (Filesystem, Size, Used, Avail, Capacity, Mounted on).

• tops-scrubber.pl -- Easy parser for top. Usage:

  perl top-scrubber.pl --type="mem" --detail="free" > output

where the type is info (line 1), tasks, cpu, mem, swap, and the detail for each line depends on the parameters parsed out. Here are some more examples.

  # 1, 5, 10-min load avg -- first row of top
  perl top-scrubber.pl --type="info" --detail="load_1"
  perl top-scrubber.pl --type="info" --detail="load_5"
  perl top-scrubber.pl --type="info" --detail="load_10"

  # idle and iowait  -- cpu row of top
  perl top-scrubber.pl --type="cpu" --detail="id"
  perl top-scrubber.pl --type="cpu" --detail="wa"

  # swap - buffers, free.
  perl top-scrubber.pl --type="swap" --detail="buffers"
  perl top-scrubber.pl --type="swap" --detail="free"

  # tasks -- running, zombie
  perl top-scrubber.pl --type="tasks" --detail="running"
  perl top-scrubber.pl --type="tasks" --detail="zombie"

Use the debug statement to get a snapshot of all values, and to see supported type flags (left col) and detail flags (indented in []'s):

  perl top-scrubber.pl --debug

• process_report.pl -- counts and averages on open processes.

Usage:

  # Reports the average cpu usage from any open processes with httpd in its command 
  perl process_report.pl --process="httpd" --metric="cpu" --summary="avg"

  # Counts how many CGI processes are open
  perl process_report.pl --process="cgi" --metric="count"

  # Reports max cpu usage of all the open foo.cgi processes
  perl process_report.pl --process="foo.cgi" --metric="cpu" --summary="max"

You can report summary= avg, max, min across metrics=cpu, mem, time (where time is the number of seconds the process has been running).