23 December 2013 Karim Elatov

I had an old box laying around that wasn’t really doing anything. It only had 1GB of RAM, but that is good enough for a splunk install.


I checked the Splunk System Requirements and I was okay:

splunk hw capacity Installing Splunk on FreeBSD

This was just for a home setup, I was only planning to monitor 6-8 machines.

Splunk Install

Looking over the download page, I saw that FreeBSD 9 is supported:

freebsd download Installing Splunk on FreeBSD

I grabbed the intel/native package and then ran the following to install it:

$ sudo pkg_add splunk-6.0-182037-freebsd-7.3-intel.tgz

The Splunk Documentation had a couple of post install instructions. From their install page:

To ensure that Splunk functions properly on FreeBSD, you must:

  1. Add the following to /boot/loader.conf

    kern.maxdsiz="2147483648" # 2GB
    kern.dfldsiz="2147483648" # 2GB
  2. Add the following to /etc/sysctl.conf:


A restart of the OS is required for the changes to effect.

If your server has less than 2 GB of memory, reduce the values accordingly.

The second setting (vm.max_proc_mmap) no longer existed on FreeBSD 9, so only I added the following to my /boot/loader.conf file:


And restarted the machine to apply the changes.

First Start for Splunk

After the reboot I decided to start the splunk service:

$ cd /opt/splunk/bin
$ sudo ./splunk start

And I saw the following:

This appears to be your first time running this version of Splunk.
Copying '/opt/splunk/etc/openldap/ldap.conf.default' to '/opt/splunk/etc/openldap/ldap.conf'.
Generating RSA private key, 1024 bit long modulus
e is 65537 (0x10001)
writing RSA key

Generating RSA private key, 1024 bit long modulus
e is 65537 (0x10001)
writing RSA key

Moving '/opt/splunk/share/splunk/search_mrsparkle/modules.new' to '/opt/splunk/share/splunk/search_mrsparkle/modules'.

Splunk> Winning the War on Error

Checking prerequisites...
WARNING: Data segment size limit (ulimit -d) is set low (1063256064 bytes)  Splunk may not work.
         You may want to run "ulimit -d unlimited" before starting splunk.
On FreeBSD the system-wide limit may need to be increased by adding the following lines to /boot/loader.conf:
  kern.maxdsiz="2147483648" # 2GB
  kern.dfldsiz="2147483648" # 2GB
:You will need to reboot for the new defaults to take effect.
        Checking http port [8000]: open
        Checking mgmt port [8089]: open
        Checking configuration...  Done.
                Creating: /opt/splunk/var/lib/splunk
                Creating: /opt/splunk/var/run/splunk
                Creating: /opt/splunk/var/run/splunk/appserver/i18n
                Creating: /opt/splunk/var/run/splunk/appserver/modules/static/css
                Creating: /opt/splunk/var/run/splunk/upload
                Creating: /opt/splunk/var/spool/splunk
                Creating: /opt/splunk/var/spool/dirmoncache
                Creating: /opt/splunk/var/lib/splunk/authDb
                Creating: /opt/splunk/var/lib/splunk/hashDb
        Checking critical directories...        Done
        Checking indexes...
                Validated: _audit _blocksignature _internal _thefishbucket history main summary
New certs have been generated in '/opt/splunk/etc/auth'.
        Checking filesystem compatibility...  Done
        Checking conf files for typos...        Done
All preliminary checks passed.

Starting splunk server daemon (splunkd)...

Starting splunkweb...  Generating certs for splunkweb server
Generating a 1024 bit RSA private key
writing new private key to 'privKeySecure.pem'
Signature ok
Getting CA Private Key
writing RSA key

If you get stuck, we're here to help.
Look for answers here: http://docs.splunk.com

The Splunk web interface is at http://moxz:8000

It did give me a warning about my ulimit settings but that’s okay since I only had 1GB of RAM. To make sure it’s running I ran the following:

$ sudo ./splunk status
splunkd is running (PID: 1151).
splunk helpers are running (PIDs: 1152).
splunkweb is running (PID: 1193).

That looked good. I went to the Splunk Web Portal and I saw the same thing as in my previous post.

Setup Splunk to Receive Syslog Logs over UDP 514

This was going to be internal, so I just wanted to setup the easiest data source. Using UDP 514 for a Syslog Server is pretty standard. By default the syslogd service running on FreeBSD listens on UDP 514. First let’s disable that, this is done by editing /etc/rc.conf file and adding the following to it:


You ran then restart the syslogd process to apply the settings:

$ sudo service syslogd restart

Now we can add a Data Source to listen on UDP 514 for Syslog Logs. From the Splunk Web Portal Home Page click on Data Inputs and you should see the following:

splunk data inputs Installing Splunk on FreeBSD

Then click on UDP and fill out all the settings:

514 data input Installing Splunk on FreeBSD

Then click Save and you should be all set. You can confirm it’s listening on that port by running the following on the Splunk server:

$ sudo /opt/splunk/bin/splunk list udp
Your session is invalid.  Please login.
Splunk username: admin
Listening for input on the following UDP ports:

Configure Machines to Log to Splunk

I only run *nix systems at home, so most of the configurations are the same. Here is what I did on my Linux systems that are running rsyslogd:

$ sudo vi /etc/rsyslog.conf

and then add the following to the bottom:

*.* @

And lastly restart the rsyslog service to apply the change:

$ sudo service rsyslog restart

On the WRT router, you can go to the Services Tab and configure it to send logs to a remote syslog server : wrt syslog Installing Splunk on FreeBSD

On OpenELEC, you can go to the OpenELEC settings and under Services you can specify a remote syslog server:

openelec remote syslog Installing Splunk on FreeBSD

After I was done with the configurations, here are all the hosts seen in splunk (under Search -> Data Summary):

data summary g Installing Splunk on FreeBSD

Setup Splunk to Automatically Start on boot

The install comes with a setup script for this. Here is what I ran to enable splunk to start on boot:

$ sudo /opt/splunk/bin/splunk enable boot-start

After that is done, you can run the following commands:

$ sudo service splunk {start,stop,restart,status}

You will also see that an entry is added to rc.conf file automatically:

$ grep splunk /etc/rc.conf

Lower Splunk Log Retention

By default 6 years worth of logs are kept. From Set a retirement and archiving policy:

Set attributes for cold to frozen rolling behavior The maxTotalDataSizeMB and frozenTimePeriodInSecs attributes in indexes.conf help determine when buckets roll from cold to frozen. These attributes are described in detail below.

More information:

You can use the size of an index to determine when data gets frozen and removed from the index. If an index grows larger than its maximum specified size, the oldest data is rolled to the frozen state.

The default maximum size for an index is 500,000MB. To change the maximum size, edit the maxTotalDataSizeMB attribute in indexes.conf. For example, to specify the maximum size as 250,000MB

And the second parameter:

You can use the age of data to determine when a bucket gets rolled to frozen. When the most recent data in a particular bucket reaches the configured age, the entire bucket is rolled.

To specify the age at which data should freeze, edit the frozenTimePeriodInSecs attribute in indexes.conf. This attribute specifies the number of seconds to elapse before data gets frozen. The default value is 188697600 seconds, or approximately 6 years.

I decided to set the size to 100GB and the time to be 90 days (7776000 seconds). So if data is more than 90 days, it’s removed. If the the index size gets bigger than 100 GB, it will be removed as well. To apply those settings edit the /opt/splunk/etc/system/default/indexes.conf file and modify the following parameters:

frozenTimePeriodInSecs = 7776000
maxTotalDataSizeMB = 100000

To apply those settings, restart the splunk server:

$ sudo service splunk restart

I don’t want to waste space on old data. I checked my Quota usage under “Settings” -> “Licensing” -> “Usage Report” and I wasn’t even close to reaching the quota:

license usage splunk g Installing Splunk on FreeBSD

Unrelated to my above settings, but still good to know.

Change the License Group to Free for Splunk

I know I have a 60 day Enterprise trial version, but before I get used to all the neat features, I wanted to setup and use the Free version of Splunk. To change the License Group, go to Settings -> Licensing:

system license g Installing Splunk on FreeBSD

Then click on Change License Group. Select the Free License and apply. It will restart splunk to apply the change. If you go back to Settings -> Licensing you should see the following:

splunk license group Installing Splunk on FreeBSD

Disable the Welcome Page for Splunk Free Version

As soon as you switch over to the Free License, you will notice that the authentication page is no longer present and splunk will check for an update and if it doesn’t find one it will ask you to continue:

splunk free version welcome screen Installing Splunk on FreeBSD

Upon clicking Continue, it will just log you without any sort of authentication. First let’s go ahead and disable that page. I ran across this splunk community page, which helped. To get rid of the welcome/update page, edit (or create if it doesn’t exist) the /opt/splunk/etc/system/local/web.conf file and add the following to it:

 updateCheckerBaseURL = 0

Then restart splunk to apply the changes:

$ sudo service splunk restart

After that when you visit the Splunk Web Portal it will just log you in without showing that page.

Configure a Reverse Proxy for Splunk Free Version

Now that there is no authentication and I wanted to check the splunk web portal externally, I decided to setup a reverse proxy for splunk. The process is described in the “Placing Splunk behind a Web proxy” page from the splunk community. I already had an apache server running on my Debian box, so I decided to utilize that server as a reverse proxy. First let’s install the mod-proxy module for apache:

$ sudo apt-get install libapache2-mod-proxy-html

Then enable the proxy modules:

$ sudo a2enmod proxy
$ sudo a2enmod proxy_http

Then setup the reverse proxy to keep the hostname of the webserver rather that showing the internal IP of the splunk server. This is done by editing the /etc/apache2/mods-enabled/proxy.conf file and adding/modifying the following:

#ProxyRequests On
ProxyPreserveHost On

Then setup the /splunk location to be proxied, this is done by editing the virtualhost configuration. In my case I only had the default one, so I edited the /etc/apache2/sites-enabled/000-default file (edit the appropriate configuration file that applies in your environment), and I added the following to it:

<Location /splunk>
  ProxyPass          retry=0
  ProxyPassReverse            retry=0

Then restart apache and you should all set:

$ sudo service apache2 restart

Change Root Endpoint for Splunk

Since we want all of our requests to start with /splunk we need to change the Splunk configuration to reflect that. This is done by editing the /opt/splunk/etc/system/local/web.conf file and making it look like this:

 root_endpoint = /splunk
 updateCheckerBaseURL = 0

To apply the settings restart splunk:

$ sudo service splunk restart

Now if you visit http://SPLUNK_SERVER:8000, you should see all the URLs start with /splunk, like so:

splunk url changed g Installing Splunk on FreeBSD

Now if you visit http://YOUR_WEBSERVER/splunk, it should take to the Splunk Web Portal.

Password Protect the Splunk Web Portal Location in Apache

Now that the reverse proxy is setup and working, let’s password protect the /splunk location. This can be accomplished with a regular htpassword file (you can even use htdigest if you want). First let’s create a directory where we will store the password file:

$ sudo mkdir /etc/apache2/pass
$ sudo chgrp www-data /etc/apache2/pass
$ sudo chmod o-rx /etc/apache2/pass

Then let’s create the htpasswd file:

$ sudo htpasswd -c /etc/apache2/pass/htpasswd admin

(I made the username be the same as the admin used, when I had the Enterprise trial)

Lastly let’s make sure only the www-data group can read that file:

$ sudo chmod 640 /etc/apache2/pass/htpasswd
$ sudo chgrp www-data /etc/apache2/pass/htpasswd

Lastly add the following section to the /etc/apache2/sites-enabled/000-default file:

  <Location /splunk>
    ProxyPass          retry=0
    ProxyPassReverse            retry=0
    AuthType Basic
    AuthName "Splunk"
    AuthUserFile /etc/apache2/pass/htpasswd
    Require valid-user

Now if you visit to your splunk instance which is behind the reverse proxy, you will be prompted for a password.

Configure FreeBSD Firewall to Only Allow Local Network Access to Splunk Web

So let’s configure the firewall to only allow TCP 8000 from the local network. This is done by editing the /etc/pf.conf file and adding the following:

# accept splunk-web sessions
pass in on $my_int proto tcp from to any port 8000 keep state

Then apply the rules:

$ sudo pfctl -f /etc/pf.conf

and make sure the rule is in place:

$ sudo pfctl -s rules | grep 8000
pass in on em0 inet proto tcp from to any port = 8000 flags S/SA keep state

That should be it. Only local access is allowed to the Splunk Web Portal directly, and to access the Splunk Web Portal externally you have to go through the reverse proxy which is password protected.

blog comments powered by Disqus