Skip to main content

Site CollectorSite Collector Administration Guide

Table of Contents

Upgrade the Site Collector Specifications

Use one of the following ways to upgrade a VM with Site Collector instance from common specifications to enterprise specifications:

Note

It is recommended to plan ahead and select the correct specification before Site Collector installation and avoid the upgrade process.

Caution

The upgrade process applies to the VMs that are created in GCP using physical ssd mounts. While a generic example is provided, failure risk exists if any step fails, hence someone with VM and Linux expertise is strongly recommended.

  • New Installation – Create a new VM with enterprise specifications and refresh the new installation to replace the old Site Collector instance.

  • Reinstallation – Delete the old Site Collector installation, update the VM configuration with enterprise specifications, then install Site Collector.

  • Manual upgrade of specifications – Use the following procedure to upgrade the specifications for a VM if new installation and reinstallation does not work.

Stop the Site Collector Instance

  1. If your Site Collector instance is not in a stopped state, use the following command to stop the Site Collector instance.

    sudo ./ngsccli sitecollector stop
    upgrade_specifications_1_updated.png

    If the Nifi queue contains a large amount of data, draining it takes time. Please wait and avoid terminating the process using Ctrl+C.

  2. Verify if the Site Collector instance has been stopped and the process is inactive by using the following command.

    systemctl status exabeam-ngsc
    upgrade_specifications_2_-_Copy.png
  3. To shut down the VM, in Google Cloud Platform (GCP), navigate to Compute Engine > Virtual machines > VM instances. Then, find and stop the VM instance.

    GCP_-_Copy.png

Upgrade the VM with Enterprise Specifications

Use the following steps to upgrade the VM with Enterprise Specifications.

  1. In the Google Cloud Platform (GCP), navigate to Compute Engine > Virtual machines > VM instances. Then, click Edit.

    upgrade_2-1.png
  2. Select a machine type with at least 16 CPU and 32GB memory and save the configuration.

    upgrade_2-2.png
  3. To update the disk size, in the GCP console, navigate to Compute Engine > Storage > Disks, and find the disks used by the VM.

    upgrade_2-3.png
  4. Edit the disk size for each of repository according to the Enterprise Specifications.

    • flowfile_repository: 75GB

    • content_repostiroy: 600GB

    • provenance_respository: 150GB

    upgrade_2-4.png
  5. To start the VM, navigate to the VM instances page, and click Start for the selected VM instance.

    upgrade_2-5_2.png
  6. SSH into the VM and inspect the the disk size by running the lsblk command in shell after the machine is up.

    upgrade_2-6.png

    If you did not configure the fstab to map the disks, re-mount the disks before running the precheck.

  7. To verify the enterprise specifications, run the following precheck for enterprise specifications using the following command and verify that the CPU and disk pre-checks are successful.

    sudo /opt/exabeam/bin/ngsccli precheck install --enterprise
    upgrade_2-7.png

Upgrade the Site Collector Instance

After verifying the CPU and disk sizes, irrespective of the Site Collector version, run the upgrade command to reconfigure the Site Collector instance to the enterprise specifications.

  1. Restart the Site Collector instance and ensure that the instance is running.

    sudo /opt/exabeam/bin/ngsccli sitecollector start
  2. Run the following upgrade command to reconfigure the Site Collector instance to utilize the increased CPU and Memory.

    sudo /opt/exabeam/bin/ngsccli sitecollector upgrade --skip-report
  3. To verify if the Site Collector instance has been upgraded to utilize the additional memory, after the upgrade, check the docker-compose config file using the following command.

    sudo less /opt/exabeam/docker-compose.yml
  4. Check the values for NIFI_JVM_HEAP_MAX and mem_limit. NIFI_JVM_HEAP_MAX should be greater than 8g and mem_limit must be greater than 11 g.

    upgrade_2-8.png

    This completes the upgrade process.