System Prerequisites for Deploying the Application Study Tool
The Application Study Tool (AST) from F5 is a powerful utility for monitoring and observing your BIG-IP ecosystem. It provides valuable insights into the performance of your BIG-IP, the applications it delivers, potential threats, and traffic patterns.
To deploy it, you’ll need a host system (physical, virtual, or cloud instance) that meets all of the prerequisites. These include machine size, network connectivity, operating system, pre-installed software packages, and user access. Whether you use dedicated hardware or provision a virtual machine, this guide will help you make sure you start with the right pre-installation configuration before you start. This will help your installation go smoothly and allow you to begin using it right away.
Note that this article is meant to provide just general guidelines for the most common installations. If you are have the tool configured to collect metrics more often than the default setting (polling every 30 seconds), or you have a unique network configuration, a highly customized Linux account permissions, or an abnormally large number of objects on your BIG-IP, your requirements will differ from the guidelines published here.
Machine Sizing
The Application Study Tool can run on most standard machine types: a dedicated hardware device, a virtual machine, or a cloud IaaS instance. However, the machine needs to have sufficient resources available for AST to run successfully.
For a small AST deployment, monitoring 20 or fewer BIG-IP devices, a Linux VM with 1-2 CPU cores and 2-4 GB RAM will suffice for typical BIG-IP usage. To monitor more devices, you should add 1 more CPU core and 2 more GB of memory for each additional 20 devices. To monitor 60 devices, you need 3-4 CPU cores and 6-8 GB of memory. For 80 devices, you need 4-5 CPU cores and 8-10 GB of memory. (These resource requirements will be higher for BIG-IPs with an abnormally high number of objects such as virtual servers, pools, policies, etc.)
Storage space will depend on how long you want to store your data. We usually suggest starting with at least 200 GB of disk space for 20 devices and 6 months of data. Double that for 12 months of history, or for 40 devices and 6 months of history. (More is always better, of course.)
Depending on the number of objects in your BIG-IP fleet and their activity, your storage needs may vary. If you increase the polling interval to collect metrics more often than the default of every 30 seconds, you will need more storage than the amount recommended here.
Operating System
Any modern Linux distribution (Ubuntu, RHEL, Rocky Linux, Debian, etc.) will work. Most customers run Ubuntu or RHEL, but we’ve also seen successful deployments on Oracle Linux and SUSE. For demo and test deployments, even macOS will work, though there are a few required tweaks beyond the scope of this guide, if you run it on an ARM-based system. (Theoretically, it should also run on Windows as long as Docker with Docker Compose is installed.)
The primary requirement is that the system supports Docker and Docker Compose (or a compatible container engine, such as Podman and Podman Compose).
Internet Connectivity
For environments with unrestricted egress access, this section should not be a concern. For those that require external sites to be explicitly whitelisted, however, you will need the ability to connect to github.com, docker.io, pypi.org, and ghcr.io -- all on port 443. (pypi.org is needed by the Configuration Helper to download the Python package before running its script.)
You can test connectivity from your AST host system by running the following to verify both connectivity and the ability to download binary content from these sites:
curl https://github.com
curl https://pypi.org
sudo docker pull grafana:latest
sudo docker pull ghcr.io/f5devcentral/application-study-tool/otel_custom_collector:latest
If you also intend to export metrics to F5, the system will need connectivity to f5.com (also on port 443). For more information on exporting metrics to F5, see the following:
Lastly, any client system (such as your desktop, a jump box, or any other users) that will be accessing the AST UI will need to be able to reach the AST host on port 3000. (By default, this is over HTTP. For encrypted communication over HTTPS, see my blog, Application Study Tool: Make Grafana Listen on HTTPS.)
For air-gapped environments and environments in which whitelisting the above sites is prohibitively difficult or forbidden by organizational policy, you can follow the instructions in the blog, How Do I Install the Application Study Tool in an Air-Gapped Environment?
User Permissions
Running the Application Study Tool should be performed by either root or a user with sudo access. However, if this is not possible, a privileged user (i.e., root) can make Docker accessible for non-root users by following these instructions: Manage Docker as a non-root user.
Software Packages
The Application Study Tool uses Git to download the required config files and runs on Docker, using Docker Compose. An alternative container engine, such as Podman and Podman Compose, will also work.
Additionally, the "Configuration Helper” can either run as a container (in which case, only Docker is needed to run it) or as a native Python script. In the latter case, Python is also required. (Note that this script just modifies the services/otel_collector/receivers.yaml file, which can be done manually if neither of the first two options are possible.) See the Run The Configuration Helper section of the installation instructions for more information on running this script.
These tools can be installed by root or a user with sudo access, if they are not already available on the system.
F5 BIG-IP Access
The last requirement is access to the BIG-IP devices you plan to monitor. The Application Study Tool will show all available metrics if configured with admin access on the BIG-IP device. However, if only a read-only account is available, the utility will still be able to show a subset of the available metrics. The Account Permissions section of the repo’s README file lists the metrics that will not be available when using a read-only account.
The full list of metrics that are collected can be found here: BigIP Receiver (AST) - Metrics
The AST host will also need to be able to connect to the BIG-IP management interfaces. These interfaces usually listen on port 443, but they can also be set up on other ports, like 8443, in some cases, like when they are set up in single-NIC mode.
Installing the Application Study Tool
Now that you have your host, network, and BIG-IP configured with the required settings, you are ready to install the Application Study Tool. Installation instructions can be found on the Application Study Overview page.
Employee