[KB6044] Upgrade ESET Remote Administrator 6 (ERA6) installed in Failover Cluster in Linux

Details

If you have ERA Server installed in a Failover Cluster environment in Linux and wish to upgrade the installation, proceed with the steps below.

Solution

Manual upgrade from version 6.3 to latest version

  1. Disable EraService in Conga (Cluster Administration GUI) under Service groups and ensure that ERA Agent and ERA Server are stopped on both nodes.
  2. Upgrade ERA Server on node1 by performing the following steps:
    • Mount the shared storage to this node.
    • Upgrade ERA Server manually to latest version by executing .sudo ./Server-Linux-x86_64.sh command in a Terminal window.
    • Replace the old cluster script located at /usr/share/cluster/eracluster_server.sh with the new one found in /opt/eset/RemoteAdministrator/Server/setup/eracluster_server. Keep the old file name.
    • Stop the ERA Server service (stop eraserver) after the upgrade
    • Disable ERA Server autostart by renaming the following 2 files:
      • mv /etc/init/eraserver.conf /etc/init/eraserver.conf.disabled
      • mv /etc/init/eraserver-xvfb.conf /etc/init/eraserver-xvfb.conf.disabled
    • Unmount the shared storage from this node
  3. Upgrade ERA Server on node2 by performing the same steps as in previous point 2.
  4. Start EraService in Conga (Cluster Administration GUI) under Service groups.
  5. Upgrade ERA Agent on all cluster nodes.
  6. Check in ERA Console if all nodes are connecting, and if Agent and Server versions for all nodes report the latest version you upgraded to.

 

Manual upgrade from version 6.1 or 6.2 to version 6.3

NOTE:

For older ERA versions the ERA Agent service was running always only on the active node in the failover cluster. Starting with ERA 6.3, the ERA Agent service is running on all nodes all the time. This way all nodes can be monitored by ERA all the time.
The change described above will cause that during upgrade new computer will be created at least for one of the cluster nodes. If you do not need event history for them, do not forget to manually remove the old computer(s) via ERA Console

  1. Disable EraService in Conga (Cluster Administration GUI) under Service groups and ensure that ERA Agent and ERA Server are stopped on both nodes.
  2. Upgrade ERA Server on node1 by performing the following steps:
    • Mount the shared storage to this node.
    • Upgrade ERA Server manually to version 6.3 by executing .sudo ./Server-Linux-x86_64.sh command in a Terminal window
    • Replace the old cluster script located at /usr/share/cluster/eracluster_server.sh with the new one found in /opt/eset/RemoteAdministrator/Server/setup/eracluster_server. Keep the old file name.
    • Stop the ERA Server service (stop eraserver) after the upgrade
    • Disable ERA Server autostart by renaming the following 2 files:
      • mv /etc/init/eraserver.conf /etc/init/eraserver.conf.disabled
      • mv /etc/init/eraserver-xvfb.conf /etc/init/eraserver-xvfb.conf.disabled
    • Unmount the shared storage from node1.
  3. Upgrade ERA Server on node2 by performing the same steps as in previous step no. 2.
  4. Remove old agent from node1 by ferforming the following steps:

    Command to list symbolic links in directory /etc/opt/:
    find /etc/opt/ -maxdepth 1 -type l -ls

    • Mount the shared storage to this node
    • Uninstall old ERA Agent (use the installer script with --uninstall parameter)
    • Check whether your system has the following symbolic links
      • /etc/opt/eset -> /usr/share/erag2cluster/etc/opt/eset
      • /opt/eset -> /usr/share/erag2cluster/opt/eset
      • /var/log/eset -> /usr/share/erag2cluster/var/log/eset
      • /var/opt/eset -> /usr/share/erag2cluster/var/opt/eset
    • If those symbolic links exist, remove them using the commands below:
      • unlink /etc/opt/eset
      • unlink /opt/eset
      • unlink /var/log/eset
      • unlink /var/opt/eset
    • Create new symbolic links - for each of them corresponding folder needs to be created first. Use the commands below:
      • mkdir -p /etc/opt/eset/RemoteAdministrator
      • ln -s /usr/share/erag2cluster/etc/opt/eset/RemoteAdministrator/Server /etc/opt/eset/RemoteAdministrator/Server
         
      • mkdir -p /opt/eset/RemoteAdministrator
      • ln -s /usr/share/erag2cluster/opt/eset/RemoteAdministrator/Server /opt/eset/RemoteAdministrator/Server
         
      • mkdir -p /var/log/eset/RemoteAdministrator
      • ln -s /usr/share/erag2cluster/var/log/eset/RemoteAdministrator/Server /var/log/eset/RemoteAdministrator/Server
         
      • mkdir -p /var/opt/eset/RemoteAdministrator
      • ln -s /usr/share/erag2cluster/var/opt/eset/RemoteAdministrator/Server /var/opt/eset/RemoteAdministrator/Server
    • Unmount the shared storage from this node
  5. Remove old ERA Agent from node2 by performing the same steps as in step no. 4.
  6. Start EraService in Conga (Cluster Administration GUI) under Service groups.
  7. Install ERA Agent on both nodes on the local disk (not on the shared cluster disk). When using the --hostname= command, you must specify the external IP address or hostname of the cluster's interface (not localhost!).
  8. Check in ERA Console if all nodes are connecting, and if Agent and Server versions for all nodes report the latest version you upgraded to.

Chat with ESET AI Advisor for support