Application-layer DDoS attacks (L7 DDoS) have been a threat since the dawn of the Internet, prompting the evolution of tools designed to combat them. Traditional open-source solutions often rely on outdated methods, such as processing text files, which can't keep up with the scale of modern attacks. Beyond L7 DDoS, various malicious bots exhaust resources through web scraping, hoarding inventory, and securing appointment slots, thus disrupting legitimate users. The rise of AI has intensified these issues, with many companies dependent on scraped data to fuel their products, leading to the creation of specialized proxy services that hide bot activity. Tempesta FW 0.8 introduces a high-performance open-source framework aimed at countering these modern threats. The upcoming version 0.9 will enhance this with tools for automated detection and blocking of sophisticated L7 DDoS attacks. A key feature is Tempesta Fingerprints, which uses a lightweight client fingerprinting mechanism that operates without the heavy computational costs associated with traditional cryptographic methods. This mechanism computes fingerprints at the TLS and HTTP layers, allowing for efficient classification of client requests. By clustering similar requests, Tempesta FW can quickly identify and filter malicious traffic, even amidst millions of connections. The article also addresses the challenges of managing access logs generated by busy servers, which can exceed 100,000 entries per second. Traditional log management becomes unwieldy due to the sheer volume of data. Tempesta FW aims to mitigate these issues by utilizing a fixed amount of memory for client accounting, efficiently tracking only the most active clients through an LRU structure. The design prioritizes performance, particularly under DDoS conditions, by leveraging 2MB huge pages for storage. This approach allows Tempesta FW to react swiftly to threats while minimizing disruption to legitimate users.