ATOP (Guide) for Linux Server Performance Analysis

We are all familiar with Apex, a real-time system monitor. Some people prefer htop and earlier, I mentioned iotop for use with storage read/write monitoring. Let’s look at another popular tool for Linux server performance analysis: Above,

benefits of above

Atop is an ASCII full-screen performance monitor that can log and report the activity of all server processes. One feature I really like is that etop will be active in the background for long server analysis (up to 28 days by default). Other benefits include:

  • Shows resource usage of all processes, even those that have been closed/completed.
  • Monitors threads within processes and ignores unused processes.
  • Stores resource usage for all processes and users with the same name.
  • Highlights important resources using colors (red).
  • Will add or remove columns as the display window is resized.
  • Includes disk I/O and network usage.
  • Use nettop Kernel module for monitoring TCP and UDP and network bandwidth.

Once atop is launched, by default, it will show system activity for CPU, memory, swap, disk, and network in 10-second intervals. In addition, for each process and thread, you can analyze CPU usage, memory consumption, disk I/O, priority, username, status, and even exit code.

Install above on RHEL/CentOS/Fedora Linux

First, install and enable EPEL (Additional Packages for Enterprise Linux) repo. See: RedHat Solution #308983,

dnf install atop

Install above on Debian/Ubuntu Linux

apt install atop

Once installed on any distro, you can launch it similar to top:

atop

Using etop – System and Process Monitor

A good place to start would be to read the man page:

man atop

Other useful commands:

Launch with average-per-second total values:

atop -1

Launch only with active processes:

atop -a

Launch with command line per process

atop -c

launch with disk info

atop -d

launch with memory info

atop -m

launch with network info

atop -n

Launch with scheduling information

atop -s

Launch with different information (ppid, user, time)

atop -v

launch with separate thread

atop -y

Once the above is running, press the following shortcut keys to sort the processes:

  • A – Sort by most active resource.
  • c – Revert to sorting by CPU consumption (default).
  • D – Sort in order of disk activity.
  • M – Sort by memory usage
  • n – sort in order of network activity

Guide to Reading Top Reports/Logs

By default the atop daemon writes snapshots to a compressed log file (eg /var/log/atop/atop_20140813) after installing. These log files can be read using:

atop -r /full/path/to/atop/log/file

Once you open a log file (for example, atop -r /var/log/atop/atop_20140813), use Tea to proceed in the interval of 10 minutes and Tea to go back. You can analyze specific time by pressing b Then enter time. The above shortcut keys also work in this mode… a, c, d, m, n,

you can use shortcut atopsa, For example, atopsar. Using flag with “-c 30 5” will prepare a report for Current CPU usage for 5 minutes (ten times with an interval of 30 seconds):

atopsar -c 30 5

Using flag -a with return all available reports.

atopsar -A

But you can limit it to a specific time window using the “-b” and end “-e” flags:

atopsar -A -b 11:00 -e 11:15

As you can see, it is easy to use Above Effectively for Linux server performance analysis.

Other popular command-line tools for Linux server performance analysis:

Above, htop, nemon, net-tools, iptraf, Collectly, vision, iostat And vmstat,

Leave a Comment