Simple steps to install & configure ClamAV in CentOS 7

ClamAV is an open source antivirus tool. Its basic usage is for detecting viruses, malware, and malicious software on Linux-based machines. The threat from viruses, Trojans, and other forms of malware is real. They have grown exponentially in both quantity and in sophistication, and antivirus software have had to adopt sophisticated detection methods. While there's no guarantee that your system will not fall victim to these unwanted bits of code, remaining mindful when using the Internet and sharing files, implementing common-sense security policies, and using an up-to-date antivirus program can go a long way in protecting you.

 

This article will show you how to install and configure ClamAV on CentOS 7 and CentOS 8. I have also added some tips for Ubuntu.

Steps to install and configure ClamAV in Linux ( CentOS / RHEL 7 )

 

1. Install ClamAV packages

To install ClamAV on CentOS 7, we need to install and enable EPEL repository.

# yum install epel-release

You can follow clamav official website to get the details of installing ClamAV on other distributions

Then we can install ClamAV with all its useful tools:

# yum -y install clamav-server clamav-data clamav-update clamav-filesystem clamav clamav-scanner-systemd clamav-devel clamav-lib clamav-server-systemd

Below is a snippet from my server after the install was successful.
Simple steps to install & configure ClamAV in CentOS 7

Below are the list of clamav rpms from my CentOS 7 environment

# rpm -qa | grep -i clamav
clamav-0.102.4-1.el7.x86_64
clamav-data-0.102.4-1.el7.noarch
clamav-filesystem-0.102.4-1.el7.noarch
clamav-update-0.102.4-1.el7.x86_64
clamav-lib-0.102.4-1.el7.x86_64
clamav-devel-0.102.4-1.el7.x86_64

 

2. Manually update the feshclam database

To update the database for the first time we need to run freshclam to update the database manually and to check whether the configuration is successfully set:

# freshclam
ClamAV update process started at Tue Nov 6 15:51:59 2018
WARNING: Can't query current.cvd.clamav.net
WARNING: Invalid DNS reply. Falling back to HTTP mode.
Reading CVD header (main.cvd): OK (IMS)
main.cvd is up to date (version: 58, sigs: 4566249, f-level: 60, builder: sigmgr)
Reading CVD header (daily.cvd): OK
Downloading daily-25006.cdiff [100%]
Downloading daily-25092.cdiff [100%]
Downloading daily-25093.cdiff [100%]
Downloading daily-25094.cdiff [100%]
Downloading daily-25095.cdiff [100%]
daily.cld updated (version: 25095, sigs: 2143057, f-level: 63, builder: neo)
Reading CVD header (bytecode.cvd): OK
bytecode.cvd is up to date (version: 327, sigs: 91, f-level: 63, builder: neo)
Database updated (6709397 signatures) from database.clamav.net (IP: 104.16.186.138)

 

This will add or update the existing database file inside

# ls -l /var/lib/clamav/
total 442156
-rw-r--r-- 1 clamupdate clamupdate    296388 Sep  5 17:16 bytecode.cvd
-rw-r--r-- 1 clamupdate clamupdate 334600704 Sep  5 14:44 daily.cld
-rw-r--r-- 1 clamupdate clamupdate 117859675 Nov 25  2019 main.cvd

 

3. Configure auto-update of freshclam database

Based on different distribution, the method to configure auto-update of freshclam database may differ. I see different behaviour in CentOS7 , CentOS 8 and Ubuntu.

 

3.1: On Ubuntu with /etc/clamav/freshclam.conf

In the /etc/clamav/freshclam.conf file of your Ubuntu machine, you'll see the following lines at the end:

# Check for new database 24 times a day
Checks 24
DatabaseMirror db.local.clamav.net
DatabaseMirror database.clamav.net

So, essentially, this means that on Ubuntu, ClamAV will be checking for updates every hour.

 

3.2: On CentOS 7 with cron job

With clamav-update-0.102.4-1.el7.x86_64 I could find a cron job file which is responsible for performing periodic update to the freshclam database

# cat /etc/cron.d/clamav-update
## Adjust this line...
MAILTO=root

## It is ok to execute it as root; freshclam drops privileges and becomes
## user 'clamupdate' as soon as possible
0  */3 * * * root /usr/share/clamav/freshclam-sleep

The */3 in the second column from the left indicates that ClamAV will check for updates every 3 hours.

 

You can change the default time to check for updates if you like, but you'll also need to change the setting in the /etc/sysconfig/freshclam file.

Let's say that you want CentOS to also check for ClamAV updates every hour. In the cron job file, change */3 to *. (You don't need to do */1 because the asterisk by itself in that position already indicates that the job will run every hour.)

Then, in the /etc/sysconfig/freshclam file, look for this line:

# FRESHCLAM_MOD=

Uncomment that line and add the number of minutes that you want between updates. To set it to 1 hour, so that it matches the cron job, it will look like this:

FRESHCLAM_MOD=60

 

To disable the auto-update you can modify

# FRESHCLAM_DELAY=

Uncomment this line and add disabled to this value:

FRESHCLAM_DELAY=disabled

 

3.3: On CentOS 8 with systemd clamav-freshclam.service

In CentOS 8 with clamav-update-0.102.4-1.el8.x86_64 I observed that below files were missing

  • /usr/share/clamav/freshclam-sleep
  • /etc/cron.d/clamav-update files
  • /etc/sysconfig/freshclam

It is possible with CentOS 8, the developer wants us to use /usr/lib/systemd/system/clamav-freshclam.service to handle auto updates of freshclam database. If you check the content of this service unit file

[Unit]
Description=ClamAV virus database updater
Documentation=man:freshclam(1) man:freshclam.conf(5) https://www.clamav.net/documents
# If user wants it run from cron, don't start the daemon.
ConditionPathExists=!/etc/cron.d/clamav-freshclam
Wants=network-online.target
After=network-online.target

[Service]
ExecStart=/usr/bin/freshclam -d --foreground=true
StandardOutput=syslog

[Install]
WantedBy=multi-user.target

We have a condition

# If user wants it run from cron, don't start the daemon.
ConditionPathExists=!/etc/cron.d/clamav-freshclam

So if /etc/cron.d/clamav-freshclam exists then user cannot start this daemon. You can find more details in this Red Hat Bugzilla where the developer seems to have done this intentionally so moving forward in RHEL/CentOS we can expect to only see the service unit file.

 

But this service unit file with CentOS 8 is not well developed to handle the auto-update of the ClamAV database.

 

With cron we had a timer which was configured to perform the auto-update. Similarly in systemd we should have an equivalent clamav-freshclam.timer file for clamav-freshclam.service but this was missing from my node.

So I decided to create my own systemd timer unit file /etc/systemd/system/clamav-freshclam.timer with below content.

# cat /etc/systemd/system/clamav-freshclam.timer
[Unit]
Description=ClamAV virus database updater
After=network-online.target

[Timer]
OnCalendar=daily
Persistent=true

[Install]
WantedBy=timers.target

 

There was one more problem though..

The existing clamav-freshclam.service is configured to start as a daemon in the foreground. When I was testing this, the timer never worked i.e. it failed to execute the freshclam daemon. The reason was because the daemon was always in running state

# ps -ef | grep freshclam
clamupd+    4874       1  0 17:14 ?        00:00:00 /usr/bin/freshclam -d --foreground=true
root        4907    2074  0 17:14 pts/1    00:00:00 grep --color=auto freshc

 

So if a daemon is already running, it is obvious that the timer won't be able to start the service again to initiate the auto update. So I decided to modify this unit file and created my own file where I am just executing freshclam without any arguments as I would do from the terminal to update the database:

# cat /etc/systemd/system/clamav-freshclam.service
[Unit]
Description=ClamAV virus database updater
Documentation=man:freshclam(1) man:freshclam.conf(5) https://www.clamav.net/documents
# If user wants it run from cron, don't start the daemon.
ConditionPathExists=!/etc/cron.d/clamav-freshclam
Wants=network-online.target
After=network-online.target

[Service]
ExecStart=/usr/bin/freshclam
StandardOutput=syslog

[Install]
WantedBy=multi-user.target

Next enable and start the clamav-freshclam.timer. We don't need to start and enable the service as timer will take care of that.

# systemctl enable clamav-freshclam.timer --now

 

So we are all done, check the status of the timer:

Steps to install and configure ClamAV in Linux CentOS 7
service status for clamav-freshclam.timer

 

Verify the list of available timers and check the time when the clamav-freshclam.timer will be next executed. So our clamav-freshclam.timer is configured to start the service next at Sun 2020-09-06 00:00:00

Steps to install and configure ClamAV in Linux CentOS 7
List the available systemd timers

 

Once the service is executed, we should see logs similar to below in journalctl

Steps to install and configure ClamAV in Linux CentOS 7
clamav database is getting updated

 

4. Configure /etc/clamd.d/scan.conf

The configuration file for ClamAV is available at /etc/clamd.d/scan.conf. The default user for performing scan is clamscan which is created as soon as we install clamav rpms

# id clamscan
uid=982(clamscan) gid=979(clamscan) groups=979(clamscan),980(virusgroup)

But we will change this to "root" user, search for

User clamscan

Comment this line and add a new line

User root

We can leave all other configuration options to default and next start the service:

 

5. Configure and start clamd.service

We have an example service file /usr/lib/systemd/system/clamd@.service that we need to copy into the system services folder.

I will copy this file to /etc/systemd/system/clamd.service. I hope you are familiar with the different systemd service file locations so you can understand why I preferred this location instead of /usr/lib/systemd/system

# cp -ap /usr/lib/systemd/system/clamd@.service /etc/systemd/system/clamd.service

 

Next replace %i with scan.conf from both the Description and ExecStart options in /etc/systemd/system/clamd.serviceSimple steps to install & configure ClamAV in CentOS 7

 

Enable and start the clamd service

# # systemctl enable clamd.service --now
Created symlink /etc/systemd/system/multi-user.target.wants/clamd.service → /etc/systemd/system/clamd.service

 

Check the status to make sure the service is active and running:

Steps to install and configure ClamAV in Linux CentOS 7
clamd service status

 

6. Configure periodic scan using clamdscan (Optional)

You can follow this step if you wish to configure auto scan of any directory as by default you will have to initiate manual scan.

We will create a new systemd service unit file :

# cat /etc/systemd/system/clamdscan-home.service
[Unit]
Description=ClamAV virus scan
Requires=clamd.service
After=clamd.service

[Service]
ExecStart=/usr/bin/clamdscan /home
StandardOutput=syslog

[Instal]
WantedBy=multi-user.target

 

To perform a periodic scan we also need a mapping timer unit file. Here I have added time value of 18:40 to start the scan:

# cat /etc/systemd/system/clamdscan-home.timer
[Unit]
Description=Scan /home directory using ClamAV

[Timer]
OnCalendar=18:40:00
Persistent=true

[Install]
WantedBy=timers.target

 

Next enable and start the timer

# systemctl enable clamdscan-home.timer --now
Created symlink /etc/systemd/system/timers.target.wants/clamdscan-home.timer → /etc/systemd/system/clamdscan-home.timer.

We don't need to start the service unit file as it will be controlled by the timer file

 

Now we monitor the journalctl logs at 18:40 PM

Sep 05 18:40:05 server.example.com systemd[1]: Started ClamAV virus scan.
Sep 05 18:40:17 server.example.com clamdscan[10901]: /home: OK
Sep 05 18:40:17 server.example.com clamdscan[10901]: ----------- SCAN SUMMARY -----------
Sep 05 18:40:17 server.example.com clamdscan[10901]: Infected files: 0
Sep 05 18:40:17 server.example.com clamdscan[10901]: Time: 11.725 sec (0 m 11 s)

 

 

7. Perform manual scan with clamscan

For a test scan of the current folder, we run the following command:

# clamscan --infected --remove --recursive ./

----------- SCAN SUMMARY -----------
Known viruses: 6702413
Engine version: 0.100.2
Scanned directories: 7
Scanned files: 9
Infected files: 0
Data scanned: 0.01 MB
Data read: 0.00 MB (ratio 2.00:1)
Time: 25.439 sec (0 m 25 s)

 

Here,

  • infected: prints only infected files
  • remove: removes infected files
  • recursive: all the sub-directories in the provided directory will also be scanned

 

Conclusion

In this tutorial we learned about ClamAV scanner and it's configuration in Linux. With different version of clamav the stepsof configuration seems to be changing. As I am more comfortable with systemd, I have used the same to demonstrate all the steps in this tutorial but you are free to write custom scripts with crond to perform auto scan and auto update of the freshclam database.

Lastly I hope the steps from the article to configure ClamAV on Ubuntu, CentOS 7 and CentOS 8 Linux was helpful. So, let me know your suggestions and feedback using the comment section.

 

References

I have used below external references for this tutorial guide
man page for systemd timer

18 thoughts on “Simple steps to install & configure ClamAV in CentOS 7”

  1. Not sure what @C.B is talking about but I just installed and configured clamav with these instructions on a Linux AWS EC2 instance with no issues...

    Excellent tutorial...
    Thanks

    Reply
  2. I also don't understand why C.B. didn't get it working. I've just installed this on CentOS7 & CentOS8 servers for mail scanning, it worked a treat on both servers. Great tutorial and clear instructions, well done. 🙂

    Reply
  3. Great tutorial! One small point, in the config file note to add the user root instead of clamscan.
    How about scanning a remote linux server? how we go about it?

    Reply
  4. thank you for spelling out the CLI switches with the double-dash option rather than just assuming everybody was born knowing what the shortcuts are. i get so irritated with articles that assume we all born knowing what "-ir" means, for example.

    Reply
  5. I used this to setup ClamAV on two CentOS 8 machines and it worked perfectly. Thank you for the guide. Very helpful.

    Reply
  6. By default clamav-update will create /etc/cron.d/clamav-update
    Make sure to disable this cron if you use the systemctl service in order to keep things clean and consistent.

    Reply
    • I am not sure if this is renamed/removed in later versions, you can try creating your own service with the provided information and if possible do let me know if it works

      Reply
  7. Good tutorial - works fine, here is proof;
    ----------- SCAN SUMMARY -----------
    Known viruses: 8351835
    Engine version: 0.102.4
    Scanned directories: 474
    Scanned files: 1552
    Infected files: 0
    Data scanned: 53.13 MB
    Data read: 25.56 MB (ratio 2.08:1)
    Time: 55.134 sec (0 m 55 s)

    Thank you !

    Reply
  8. 1) You have first

    # mv /usr/lib/systemd/system/clamd@.service /usr/lib/systemd/system/clamd.service

    and then you want to edit the file which was just renamed

    # vim /usr/lib/systemd/system/clamd@scan.service

    IMHO John above probably refers to that ...

    2) You rename and/or edit files in /usr/lib/systemd/system (and/or similar). These belong to the installed packages (check with `rpm -qf`) and removing/changing them is a total no-go (unless you never will simply `dnf update` it).
    For starters: With systemd, the files can be overloaded with a file with the same name somewhere under /etc/systemd/ (or disabled/ignored with a symlinkl there to /dev/null).

    Reply
    • Hi Bernd,

      Thank you for highlighting this, I may have overlooked earlier. I agree we should prefer to use /etc/systemd/system but since we are not using the default clamd@scan.service for our configuration, a dnf update will not break anything.

      Based on your feedback I did some trials today and realised alot has changed since I wrote this article, so I have updated the article completely based on the new clamav packages

      Reply
  9. If you have SELinux set to enforcing you will have issues when attempting to start the clamd service. This was the only error I could see from systemctland journalctl

    Sep 22 09:28:40 localhost.localdomain systemd[1]: clamd.service: Failed with result 'exit-code'.

    However once I tried to enable logging I received a new message through journalctl

    Sep 22 10:14:41 localhost.localdomain python3[27818]: SELinux is preventing clamd from open access on the file /var/log/clamd.scan

    I set SELinux to permissive, and disabled it in the configuration file.

    [root@localhost ~]# setenforce Permissive
    /etc/selinux/config
    SELINUX=disabled

    Now clamd.service was no longer having any issues starting.
    A note for Fedora, CentOS, and Redhat users.

    Reply

Leave a Comment

Please use shortcodes <pre class=comments>your code</pre> for syntax highlighting when adding code.