Logs in Linux are stored in the /var/log directory.

Remember that /var directory stands for variable, meaning,

files that constantly change are kept in this directory,

and it turns out that logs are constantly changing.

If we look at the /var/log directory with an LS,

it might seem a little intimidating. Don't worry.

Each of these log files store

specific information that we can figure out by their file names.

Let's check out some of the common ones you'll look at,

/var/log/auth.log, authorization and security-related events are logged here.

/var/log/kern.log, kernel messages are logged here.

/var/log/dmesg, system startup messages are logged here.

If you encounter an issue at, let's say,

boot up, this is a good place to check for information.

It might get a little tiresome to open up each of

these log files to find information about events.

Luckily, there are also log files that combine the information of other log files.

The downside is that these files are usually very large.

If you have a pretty good idea of where a problem might lie,

you might want to opt for the smaller and more specific log file.

The one log file that logs pretty much everything on

your system is a /var/log/syslog file.

The only thing that sys log doesn't log by default are off events.

When troubleshooting issues with user machines,

/var/log/syslog will usually contain

the most comprehensive information about your system,

so that should be your first stop.

Log files output a lot of events.

By that logic, they take up a lot of data that has to be stored on our machine somewhere.

We generally just want to see the latest events on our system,

so we don't need to overload or disk with all this information.

Luckily, our systems also do a good job of

cleaning out log files to make room for new ones.

They use something called log rotation to do this.

In Linux, the utility rotate logs is called log rotate.

You might want to investigate an event that happened a month ago,

so you can change your log rotation settings to

make sure not to delete events that are that old.

We won't discuss how to work with log rotation,

but you can read more about it in the supplemental reading.

We've talked about logging in the context of a single machine,

but if you find yourself managing many systems and want

to be able to parse their logs in one central location,

you can use something called centralized logging.

We won't talk about how to do this,

but if you're interested in setting up a centralized server,

check out the next supplemental reading.

Okay, enough talk about what logs are.

Let's actually look at some real ones.

Whoa, this looks super intimidating.

But don't worry, we're not going to be reading all of this.

In the next lesson, we'll teach you how to troubleshoot using logs.

But for now, let's just read one line in sys log and parse what it says.

The first field here is the time stamp when the event occurred. Pretty straightforward.

But depending on the log,

you might see a time format you aren't familiar with like a long string of

numbers such as 1501538594.

Time stamps found in a format like this are referred to as Unix or epoch time.

At first, you might be baffled by this.

Why would you represent time in this way?

And just what exactly is the Unix epoch?

The Unix epoch time is used to represent, then,

it's the number of seconds since midnight on January first,

1970, a sort of zero hour for Unix based computers to anchor their concept of time.

This means that 1501538594 represents the date,

time, Monday, July 31st.

30314 Pacific Standard Time.

Why midnight on January first, 1970?

Is that date the birthday of Unix?

Or does it mark some other significant event?

The actual answer is much simpler.

The original engineers of Unix at Bell Labs just picked it because it was convenient.

So, don't be caught off guard if you see a time stamp like this.

The next field is the host name of the machine the event occurred on.

Next up is the service that the log event is referring to.

And last is the event that occurred.

In the next lesson, we'll show you some common troubleshooting tactics when using logs.

Linux Logs

Operating Systems and You: Becoming a Power User

Conheça os instrutores