How to use SAR on Redhat

Eyeball Sar is a system monitor command used to display system activity. Sar ts installed via the sysstat rpm. Use the command below to install sysstat.

[root@fedora ~]# yum -y install sysstat

Once installed you can check out the sysstat config file (/etc/sysconfig/sysstat) and configure how long to sar will keep your logfiles, on my system the default was 7 days. I changed this to 30 days.

The cron job for sar is located here (/etc/cron.d/sysstat) if you want to modify it as well.

Once installed and configured to your liking you must ensure that it starts and runs at boot time. I accomplished this via the command below.

[root@fedora ~]# chkconfig sysstat on && service sysstat start

Once up and running it will write its logs out to /var/log/sa, and you can read those files with the following command (sar -d <filename>, where filename is the name of the file that you want to read). Note that in the examples below, 3 is the interval, and 10 is the count.

Additionally you may run sar interactively, below are a few sample commands.

View disk i/o and transfer rate stats:

sar -b 3 10

View memory and swap space stats:

sar -r 3 10

View swapping stats:

sar -W 3 10

View network stats:

sar -n DEV 3 10

View CPU stats:

sar -P ALL 3 10

One thought on “How to use SAR on Redhat

  1. A minor thing – when you install sysstat package, you also add /etc/cron.d/sysstat file, which samples performance matrices once every 10 minutes, by default, and generates a simple text file (which you can read with kSar, if you’re not into parsing-by-your-eyes mode). The problem with 10 minutes averages, is that they say very little. Their stats are so normalized you can hardly do anything with that.
    What I do is modify the sample time from */10 (once every 10 minutes) to */1 (once every minute). I also add the ‘-d’ flag to sa1, as it does not sample disk performance by defaults (starting at RHEL5 and above. I think this is the case for RHEL6, but I did not play much with it). Starting at that moment, you will have better statistics. You can post-mortem some of your problems after they are detected, and see if you can figure out what was the cause.
    Anyhow, I enjoyed your blog. Might be back once a while 🙂


Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s