Upgrade a distributed AMI or OVA installation of Splunk UBA

Perform the following steps to upgrade a distributed AMI or OVA installation of Splunk UBA.

Before you begin

Before upgrading Splunk UBA in a distributed deployment, confirm you meet the Upgrade Splunk UBA prerequisites. Make sure that the prerequisites are verified on each server in the distributed deployment.

If you have enabled the integration that sends UBA audit data to Splunk Enterprise Security (ES), preserve the associated certificate before upgrading. For more info on sending audit data to ES see, Send Splunk UBA audit events to Splunk ES. Use the following command to export the certificate:

. /opt/caspida/bin/CaspidaCommonEnv.sh
sudo keytool -exportcert -alias "splunk es" -keystore $JAVA_HOME/jre/lib/security/cacerts -rfc -file ~/splunk-es_cacert.pem

If you have enabled validation of the SSL certificate from Splunk ES, export the SSL certificate used for validating datasources from the Splunk Enterprise platform:

. /opt/caspida/bin/CaspidaCommonEnv.sh
sudo keytool -exportcert -alias "SplunkESRootCA" -keystore $JAVA_HOME/jre/lib/security/cacerts -rfc -file ~/SplunkESRootCA.pem

Distributed environment upgrade steps

The Splunk UBA AMI and OVA images use Ubuntu as the operating system. Ensure that Splunk UBA is running before you upgrade, then perform the following tasks to upgrade Splunk UBA across multiple servers with the Ubuntu operating system.

To obtain and install Splunk UBA 5.1.0, perform the following tasks on the management node:

1. On the management node only, stop all the caspida services.

   /opt/caspida/bin/Caspida stop-all

2. On each node, archive the previous UBA release. Replace <old-uba-version> with your version of UBA:

   sudo mkdir -p /var/vcap/release_archives
   sudo mv /opt/caspida /var/vcap/release_archives/caspida-<old-uba-version>
   sudo mkdir -p /opt/caspida && sudo chown caspida:caspida /opt/caspida
   sudo chown caspida:caspida /var/vcap/release_archives
   sudo chown caspida:caspida /var/vcap/release_archives/caspida-<old-uba-version>

.
3. On each node, download the latest UBA branch build splunk-uba-software-upgrade-package_510.tgz and untar the UBA bits to /home/caspida directory:

   $ tar xvzf /home/caspida/splunk-uba-software-upgrade-package_510.tgz
   Splunk-UBA-Platform-5.1.0-20220721-000019.tgz
   Splunk-UBA-Platform-5.1.0-20220721-000019.tgz.md5sum
   uba-ext-pkgs-5.1.0.tgz
   uba-ext-pkgs-5.1.0.tgz.md5sum
   

4. On each node, untar the UBA 5.1.0 Platform build inside the /opt/caspida/ folder:

   tar xvzf /home/caspida/Splunk-UBA-Platform-5.1.0-20220721-000019.tgz -C /opt/caspida

5. Update the list of allowed serialization classes:

   sed -i '${s/$/,com.caspida.clients.jobmanager.JobDataParams$SourceTypes,com.caspida.clients.jobmanager.JobDataParams$SourceFiles/}' /opt/caspida/conf/uba-default.properties

6. On each node, untar the external packages in the home folder.

   tar xvzf /home/caspida/uba-ext-pkgs-5.1.0.tgz -C /home/caspida

7. On each node, remove grub-pc package, then reinstall. This is done because there is an existing Ubuntu upgrade bug that is avoided with a clean install of this package.
   1. Remove by purging the grub-pc package:

      sudo apt-get purge grub-pc -y

   2. Reinstall grub-pc package:

      sudo apt-get install grub-pc -y 

   3. When the interactive prompt comes up to choose a disk, select your BOOT disk if known. If BOOT disk is not known, select all disk choices.
      

      Failure to choose the correct disk will result in it not being able to boot up.

8. On each node, reboot the system:

   sudo reboot

9. On each node, upgrade the OS from Ubuntu 16.04 to Ubuntu 18.04. UBA 5.1.0 is Ubuntu 18.04 minimum.

   tmux
   sudo /opt/caspida/upgrade/utils/upgrade_ubuntu.sh /home/caspida/uba-ext-pkgs-5.1.0

10. On each node, after Ubuntu is upgraded, reboot the system:

    sudo reboot

11. On the management node, run the UBA upgrade script, using the path to your archived UBA from step 2. Migrate your configurations in a tmux shell just in case your SSH connection breaks.

    The path-to-prev-uba-archive might be /var/vcap/release_archives/caspida- depending on if your archived UBA version is 5.0.5 or 5.0.5.1.

    $ tmux
    $ /opt/caspida/upgrade/utils/upgrade_5.0_to_5.1.sh -p <path-to-prev-uba-archive> -e /home/caspida/uba-ext-pkgs-5.1.0.tgz

    The command installs the new Splunk UBA software, restarts Splunk UBA, and then restarts the data sources.
12. Perform these steps to finalize the hadoop upgrade:
1. On the UBA management node run the following commands:

   sudo -u hdfs hdfs dfsadmin -finalizeUpgrade

   sudo service hadoop-hdfs-namenode restart

2. Restart UBA:

   /opt/caspida/bin/Caspida stop-all

   /opt/caspida/bin/Caspida start-all

If you have previously imported an output connector certificate, re-import the certificate. See, Configure the Splunk platform to receive data from the Splunk UBA output connector in the Send and Receive Data from the Splunk Platform manual.

Re-import the certificate used for sending UBA audit events to Splunk ES

If you have enabled the integration that sends UBA audit data to Splunk Enterprise Security (ES), and you preserved the associated certificate in the Before you begin step, you can now re-import that certificate. Use the following command to re-import the certificate:

. /opt/caspida/bin/CaspidaCommonEnv.sh
sudo keytool -import -alias "splunk es" -keystore $JAVA_HOME/jre/lib/security/cacerts -rfc -file ~/splunk-es_cacert.pem

Re-import the SSL certificate used for validating datasources from the Splunk Enterprise platform

If you have enabled the integration that sends UBA audit data to Splunk Enterprise Security (ES), use the following command to re-import the SSL certificate:

. /opt/caspida/bin/CaspidaCommonEnv.sh
sudo keytool -import -alias "SplunkESRootCA" -keystore $JAVA_HOME/jre/lib/security/cacerts -file ~/SplunkESRootCA.pem

For more information on SSL certificate validation, see Configure flag to enable or disable Splunk SSL certificate validation.

Apply security patches on Ubuntu

The Splunk UBA AMI and OVA images use Ubuntu as the operating system. Perform the following tasks to apply the latest security patches to your Ubuntu operating system:

Applying the security patches can take up to one hour.

1. Log in to the Splunk UBA management node as the caspida user.
2. On the management node, run the following command to stop Splunk UBA and all services:

   /opt/caspida/bin/Caspida stop-all

3. On each Splunk UBA node, perform the following tasks:
   1. Log in as the caspida user.
   2. Run the following commands to install latest unattended-upgrades package:

      sudo apt-key adv --keyserver hkp://keyserver.ubuntu.com:80 --recv-keys 6A030B21BA07F4FB
      sudo apt update
      sudo apt install unattended-upgrades
      

      If you see the following prompt, select keep the local version currently installed:

      What do you want to do about modified configuration file 50unattended-upgrades?

   3. Run the following command:

      sudo apt autoremove

   4. Edit the /etc/apt/apt.conf.d/50unattended-upgrades file and un-comment the following line:

      Skip this step if you have previously applied security patches to your Ubuntu environment following these instructions.

      "${distro_id}:${distro_codename}-security";

      Leave all other lines commented out.
   5. Run the following command:

      sudo unattended-upgrade -d

   6. Edit the /etc/init.d/zookeeper-server file and change su to runuser in all of the following lines:

      Skip this step if you have previously applied security patches to your Ubuntu environment following these instructions.

      Before:

      su -s /bin/bash zookeeper -c "${DAEMON_SCRIPT} start"
      su -s /bin/bash zookeeper -c "${DAEMON_SCRIPT} stop"
      su -s /bin/bash zookeeper -c "zookeeper-server-initialize $*"
      

      After:

      runuser -s /bin/bash zookeeper -c "${DAEMON_SCRIPT} start"
      runuser -s /bin/bash zookeeper -c "${DAEMON_SCRIPT} stop"
      runuser -s /bin/bash zookeeper -c "zookeeper-server-initialize $*"
      

   7. Reboot the system:

      sudo reboot

4. On the management node, run the following command to start Splunk UBA and all services:

   /opt/caspida/bin/Caspida start-all

Troubleshoot the Ubuntu upgrade

While upgrading Ubuntu some environment specific issues can arise. See the following known issues and how to work around them.

Snapd

You might encounter the following error message if snapd is in an unexpected state:

FileNotFoundError: [Errno 2] No such file or directory: 'snap'

Workaround steps:

1. Remove and purge the snapd package:
   sudo apt purge snapd
2. Reinstall the snapd package:
   sudo apt install snapd
3. Re-run the upgrade_ubuntu.sh script.

Zookeeper server missing

You might encounter the following error if zookeeper is upgraded to a newer version, causing the removal of the zookeeper-server: Can't select installed nor candidate version from package 'zookeeper-server' as it has neither of them

Workaround steps:
Skip holding this package.

1. Open and edit /opt/caspida/upgrade/utils/upgrade_ubuntu.sh using a text editor.
2. Modify line 177 from the following:
   CMD="apt-mark hold zookeeper zookeeper-server redis-server redis-tools"
   To appear as follows:
   CMD="apt-mark hold redis-server redis-tools"
3. As applicable, repeat for all nodes in your cluster.
4. Re-run the upgrade_ubuntu.sh script.

Next steps

Verify a successful upgrade of Splunk UBA.