Keepalived is a production-grade Linux daemon for high-availability and load balancing. It implements VRRP v2/v3 for router failover, integrates with the Linux Virtual Server (IPVS) kernel module for Layer 4 load balancing, and includes a suite of health checkers (HTTP, SSL, TCP, UDP, SMTP, DNS, PING, MISC) plus BFD for fast failure detection. The architecture is multi-process: a parent process manages dedicated children for VRRP, checkers, and BFD, coordinating reloads and termination via signals and eventfd. It deeply interacts with the Linux kernel through netlink, nftables/iptables, raw sockets, and proc connector, and exposes operational interfaces via SNMP, D-Bus, JSON state dumps, SMTP alerts, and notification scripts. The codebase is written in C with careful attention to Linux-specific features, including network namespaces, VRFs, VMAC/macvlan, and multicast group handling. It is a mature, widely deployed critical infrastructure component rather than a research artifact or library.
This is a large community-maintained repository of Zabbix monitoring templates. It provides YAML/XML template definitions and auxiliary scripts for monitoring a vast array of IT infrastructure, including network devices, servers, applications, printers, UPS, cloud services, and more. The repository is organized by category and Zabbix version (5.0–7.4). Each template typically contains item definitions, triggers, discovery rules, and macros. Data collection methods range from SNMP, IPMI, Zabbix agent, HTTP agent, to external scripts. The codebase is not a standalone application but a library of configurations consumed by Zabbix administrators. Quality is inconsistent: many templates lack triggers, have hardcoded defaults, or are incomplete. The repository includes a CI pipeline for validating template structure, naming, and import compatibility across multiple Zabbix versions. While some templates exhibit good patterns (e.g., low-level discovery, dependent items for JSON parsing), most are straightforward and tightly coupled to Zabbix internals, limiting reusability outside the Zabbix ecosystem. Overall, the repo is valuable for Zabbix users needing ready-made monitoring, but it suffers from community contribution variability, version duplication, and documentation gaps. It is actively maintained but individual templates may be stale.