Configure Linux file system monitoring

By default, Elastic Defend monitors specific Linux file system types that Elastic has tested for compatibility. If your network includes nonstandard, proprietary, or otherwise unrecognized Linux file systems, you can configure the integration policy to extend monitoring and protections to those additional file systems. You can also have Elastic Defend ignore unrecognized file system types if they don’t require monitoring or cause unexpected problems.

Ignoring file systems can create gaps in your security coverage. Use additional security layers for any file systems ignored by Elastic Defend.

To monitor or ignore additional file systems, configure the following advanced settings related to fanotify, a Linux feature that monitors file system events. Go to Manage → Policies, click a policy’s name, then scroll down and select Show advanced settings.

Even when configured to monitor all file systems (ignore_unknown_filesystems is false), Elastic Defend will still ignore specific file systems that Elastic has internally identified as incompatible. The following settings apply to any other file systems.

linux.advanced.fanotify.ignore_unknown_filesystems

Determines whether to ignore unrecognized file systems. Enter one of the following:

true: (Default) Monitor only Elastic-tested file systems, and ignore all others. You can still monitor or ignore specific file systems with monitored_filesystems and ignored_filesystems, respectively.
false: Monitor all file systems. You can still ignore specific file systems with ignored_filesystems.

If you’ve upgraded from 8.3 or earlier, this value will be false for backwards compatibility. If you don’t need to monitor additional file systems, it’s recommended to change ignore_unknown_filesystems to true after upgrading.

linux.advanced.fanotify.monitored_filesystems

Specifies additional file systems to monitor. Enter a comma-separated list of file system names as they appear in /proc/filesystems (for example: jfs,ufs,ramfs).

It’s recommended to avoid monitoring network-backed file systems.

This setting isn’t recognized if ignore_unknown_filesystems is false, since that would mean you’re already monitoring all file systems.

Entries in this setting are overridden by entries in ignored_filesystems.

linux.advanced.fanotify.ignored_filesystems

Specifies additional file systems to ignore. Enter a comma-separated list of file system names as they appear in /proc/filesystems (for example: ext4,tmpfs).

Entries in this setting override entries in monitored_filesystems.

Find file system names

edit

This section provides a few ways to determine the file system names needed for linux.advanced.fanotify.monitored_filesystems and linux.advanced.fanotify.ignored_filesystems.

In a typical setup, when you install Elastic Agent, Filebeat is installed alongside Elastic Endpoint and will automatically ship Elastic Endpoint logs to Elasticsearch. Elastic Endpoint will generate a log message about the file that was scanned when an event occurs.

To find the system file name:

From the Hosts page (Explore → Hosts), search for message: "Current sync path" to reveal the file path.
If you have access to the endpoint, run findmnt -o FSTYPE -T <file path> to return the file system. For example:
```
> findmnt -o FSTYPE -T /etc/passwd
FSTYPE
ext4
```
This returns the file system name as ext4.

Alternatively, you can also find the file system name by correlating data from two other log messages:

Search the logs for message: "Current fdinfo" to reveal the mnt_id value of the file path. In this example, the mnt_id value is 29:
```
pos:	12288
flags:	02500002
mnt_id:	29
ino:	2367737
```
Search the logs for message: "Current mountinfo" to reveal the file system that corresponds to the mnt_id value you found in the previous step:
```
<snip>
29 1 8:2 / / rw,relatime shared:1 - ext4 /dev/sda2 rw,errors=remount-ro
<snip>
```
The first number, 29, is the mnt_id, and the first field after the hyphen (-) is the file system name, ext4.

« Configure self-healing rollback for Windows endpoints Create an Elastic Defend policy using API »